CDS-bench: Task definitions

84 tasks total: 65 individual task types and 19 composition tasks.

Proposal for evaluating procedural reasoning and constrained design on coding DNA sequences. This document specifies task types, easy and hard prompt templates, scoring rules, and composition tasks.

Draft proposal · Version 0.3 · 26 October 2026

Six examples from the 84 tasks

Four individual tasks and two composition tasks. Each starts from supplied coding DNA and leads to one required answer. For RNA and protein tasks, input preparation is labeled separately. Task titles link to the full definitions.

1 · Individual task

Translate DNA to protein

CDSATG GCT TAA

Read consecutive codons · NCBI code 11

ATGM

GCTA

TAASTOP

Final answer · scalar

MA

The first in-frame stop ends translation and is omitted from the answer.

2 · Individual task

Disrupt a specified local hairpin

CDS · RNA view uses T → UATG GGA AAA TCC TAA

Expose/edit bases 10–11 only; preserve MGKS.

Supplied RNA stem3 canonical contacts

After TCC → AGC1 canonical contact

Final answer · scalar

ATGGGAAAAAGCTAA

Two targeted stem contacts break; protein MGKS remains unchanged.

3 · Individual task

Digest with trypsin rules

CDSATG GCT AAA CGT CCT GCT AAA TAA

Input preparation: translate to MAKRPAK.

Cut after K/R, except before P

1M2A3K4R5P6A7K

K3 │ cut · R4–P5 stays intact

Final answer · json

[
 {"start":1,"end":3,"peptide":"MAK"},
 {"start":4,"end":7,"peptide":"RPAK"}
]

Return both fragments in protein order, with inclusive residue coordinates.

4 · Individual task

Remove specified DNA motifs

CDSATG GAA TTC TGA

Remove GAATTC; preserve protein MEF and fixed terminal codons.

Search synonymous encodings

Phenylalanine codonTTC → TTT

Amino acidF → F

Rewritten CDSATG GAA TTT TGA

GAATTC absent · MEF unchanged

Final answer · scalar

ATGGAATTTTGA

Choose the first valid DNA string in A<C<G<T order; a valid rewrite alone is insufficient.

5 · Composition task

From CDS panel to a diagnostic peptide

Target CDS

Comparison CDS

ATG GCT AAA GCT GGT AAA TAAATG GCT AAA GGT GCT AAA TAA

Translate both

MAKAGKMAKGAK

Complete trypsin digestion

MAK │ AGKMAK │ GAK

Discard shared peptides; require length 3

Final answer · scalar

AGK

MAK is shared. AGK occurs only in the target digest; select by length, then alphabetical order.

6 · Composition task

From specified RNA edits to protein

CDSATG CAA GGT TAA

Transcribe

Coding RNAAUG CAA GGU UAA

Supplied edit: position 4, C → U

Edited RNAAUG UAA GGU UAA

Translate from the original start

Final answer · scalar

The new UAA stops translation before GGU. The unedited CDS would encode MQG.

Biological scope and supplied inputs

The starting sequence is a complete, spliced CDS with its coding strand and translation start supplied. Tasks do not require gene discovery, splice-site identification, or recognition of a protein family from learned sequence patterns.

What an instance can supply

A CDS, or a clearly identified panel of CDSs.
A requested edit, motif, guide, pairing map, or coding module.
Custom score tables, allowed edits, and numerical bounds.

What the benchmark does not ask for

Gene discovery, arbitrary genomic splicing, or isoform prediction.
Unannotated biological function, folding, expression, or guide efficacy.
Noncoding-RNA maturation or whole-transcript processing.

RNA and protein inputs are obtained by transcribing or translating the supplied CDS. Composition tasks explicitly combine the named operations and specify the final answer.

Shared sequence conventions

5′→3′ coding DNA; bases may be separated by spaces. All coordinates are 1-based inclusive unless a field explicitly says otherwise. Ordinary translation uses NCBI code 11. First in-frame stop terminates; no stop character appears in the protein.

Synonymous rewrites keep length, first codon, terminal stop, and translated protein fixed. Alternative codes can be separate future rule sets; v0.3 does not silently substitute one.

Easy and hard variants

Each instance has an easy prompt and a hard prompt. They differ only in the biological reference information supplied. The sequence, edits, constraints, correct answer, and scoring rules are the same in both. When the task rules already supply everything needed, the two prompts are identical.

Example: Calculate mean hydropathy (protein.mean_hydropathy). Both variants ask for the average hydropathy of the protein encoded by the same CDS. These illustrative excerpts abbreviate the reference tables and omit the shared answer-format instructions.

variant = easy

Reference supplied

The prompt supplies translation rules, all 64 codon assignments, and all 20 amino-acid hydropathy scores.

REFERENCE
Read the CDS in groups of three bases from base 1. The annotated initiator gives M. End at the first in-frame stop; omit it from the protein.

NCBI code 11 codons:
ATG=M, GCT=A, TAA=STOP, … [64 entries]

Kyte–Doolittle residue scores:
A=1.8, M=1.9, V=4.2, … [20 entries]

TASK
Translate the CDS and average the scores of its amino acids. Return one number with 4 decimal places.

cds: {{cds}}

variant = hard

Reference knowledge assumed

The model must know the codon assignments, initiator and stop rules, and numerical residue scores implied by the two reference names.

Translate this CDS using NCBI genetic code 11. Return its whole-protein mean Kyte–Doolittle hydropathy as one number with 4 decimal places.

cds: {{cds}}

A custom pairing-score table, edit budget, allowed mutation set, or host-specific codon weights must remain supplied in both variants. Those are instance data. Hard prompts should never require guessing an unpublished convention.

A hard prompt can name an established biological reference and expect the model to supply the underlying facts. For example, “use the Kyte–Doolittle scale” requires knowing the hydropathy score for each amino acid. “Translate using NCBI genetic code 11” requires knowing the codon-to-amino-acid assignments, stop codons, and how annotated initiator codons are translated. Easy prompts provide these tables and rules explicitly; hard prompts test whether the model already knows them.

Task schema and scoring

Each task defines its inputs, answer format, and one canonical answer. An instance supplies the actual sequences and conditions. All 84 tasks use the same scorer.

Individual task type: task_type names one of 65 individual operations. For example, cds.translate asks for the protein encoded by a CDS.
Composition task type: composition_type names one of 19 tasks that combine individual operations. For example, pipeline.motif_to_dna translates a CDS, finds a peptide motif, then maps its location back to DNA. Use this field instead of task_type for a composition task.
Variant: variant is easy or hard. It changes biological reference support; both variants share the answer format and correct answer.
Answer format: output_format is scalar for a direct value or simple list, or json for maps and records. output_rules defines every value’s meaning, ordering, precision, and failure form. Prompts show a compact output structure with typed placeholders. The full output_schema remains in the task metadata for automated validation.
Unique reference: objective and selection_rule determine one reference answer, including all tie-breaks. Reference solvers verify constraints, optimality, and any claim of impossibility before evaluation.
Scorer: scorer = exact_match for individual and composition tasks alike. It compares the complete parsed answer with the verified reference.

Each identifier has the form prefix.operation. The prefix indicates the task family or how component tasks are combined. These summaries explain the prefixes; the individual task catalogue and composition catalogue define the questions.

task_type · individual task families

Individual task prefix meanings
Prefix	Meaning
`dna`	Operations on DNA bases, strand orientation, transcription, and DNA-pattern matches.
`cds`	Interpret complete coding sequences, including translation, validity, and equivalence of encoded proteins.
`coords`	Convert positions between DNA bases, aligned codons, and protein residues.
`variant`	Apply specified DNA edits or determine their effects on the reading frame and protein product.
`ambiguity`	Resolve possible codon meanings when observed DNA contains uncertain bases represented by IUPAC symbols.
`stats`	Measure nucleotide composition and codon usage, across a whole CDS or within local windows.
`protein`	Analyze the protein encoded by a CDS, including sequence patterns, physical properties, and cleavage products.
`rna`	Analyze CDS-derived coding RNA, including edits, translation, complementarity, secondary structure, and cleavage products.
`design`	Rewrite coding DNA under specified constraints while preserving its encoded protein.
`construct`	Create or process engineered coding constructs, including protein changes, fused modules, tags, deletions, and cleavage products.
`library`	Analyze or design mixtures of codons represented by IUPAC symbols. Here ambiguity specifies a designed mixture, rather than uncertainty about one observed sequence.
`assay`	Select CDS-based primers or probes under stated sequence-matching, specificity, and uniqueness rules.
`overlap`	Analyze or modify DNA read in two explicitly annotated coding windows, accounting for both encoded products.

composition_type · pipeline and joint

Composition prefix meanings
Prefix	Meaning
`pipeline`	Several operations connected by their results. An earlier step supplies the sequence, annotation, or measurement used by later steps. Workflows can branch and can include design followed by verification. Example: Edit DNA → translate → digest.
`joint`	One design evaluated against several requirements at once. The same candidate CDS must meet every constraint and any stated optimization objective; solving the component requirements separately is insufficient. Example: Remove restriction sites while enforcing GC bounds.

Evaluating a response

Read and validate. The answer starts on the line after a standalone Final answer: marker and continues to the end of the response. JSON may span multiple lines. Parse the declared format and validate its fields and types against the task metadata.
Compare the contents. Apply only the declared formatting normalizations. Compare JSON objects by their keys and values, and lists in their required order. Every requested field must match, including counts and rounded numbers.

Score 1 only when both steps pass; otherwise 0. Report format compliance separately from answer accuracy. Composition tasks receive one score for the complete answer.

For example, the codon-library task expects:

{"status":"ok","codon":"GAK","residue_counts":{"D":1,"E":1}}

Reordering object keys is harmless. Changing a count, omitting a field, or choosing a different codon fails.

Shared answer conventions

An optional explanation precedes the marker. The answer itself begins on the next line and is the final part of the response. JSON is one complete object or array; indentation and line breaks are optional. Ignore surrounding whitespace and declared sequence spacing. Object-key order and equivalent numeric spellings such as 1.8500 and 1.85 do not matter. Preserve list order, multiplicities, identifiers, and values. Malformed JSON, duplicate keys, unexpected fields, and changed sequences fail validation.

Round once at the task’s specified precision, with exact halfway cases rounded away from zero; rounded negative zero becomes zero. Use declared reference constants and unrounded intermediate values. CAI optimization on a fixed-length CDS compares exact products of supplied decimal weights before taking roots or rounding.

Empty lists are []. Empty standalone sequences use EMPTY_SEQUENCE; empty sequences inside JSON use "". NONE, INVALID, and IMPOSSIBLE have the meanings defined by their tasks. JSON tasks specify their exact failure objects and use null where required.

Individual tasks (65)

Each row states what is given and what the model must do. Select a row for a task outline followed by easy and hard example prompts with complete inputs and computed answers. Format and scoring details follow the examples. Data provenance appears below each expected answer. Most examples reuse the bacterial CDS documented in Example data sources; engineered changes are identified per example. Expand Template for the unfilled version. The copy icon copies its code block verbatim. Search or filter the table to compare related tasks.

Example data sources

Most examples reuse the complete Escherichia coli K-12 MG1655 trpL CDS, a 45-base gene encoding a 14-residue leader peptide. Source: NCBI Gene 945856; locus b1265; nucleotide accession NC_000913.3, bases 1,323,038–1,323,082 on the minus strand; protein accession NP_415781.1. Retrieved 26 October 2026.

Native CDS, coding strand 5′→3′
ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA
Protein: MKAIFVLKGWWRTS

Engineered minigene: native codons 1–7 + native stop
ATG AAA GCA ATT TTC GTA CTG TGA
Protein: MKAIFVL

The truncated minigene keeps exact synonymous searches small enough to verify exhaustively. Each example identifies any further mutations, ambiguous codons, fusion modules, or comparison variants in its Data provenance note. Supplied RNA contacts and edits are exercises, not claims about a natural trpL structure or editing event. Codon weights are illustrative instance data, not measured expression estimates.

01 /

CDS interpretation

Orientation, translation, coordinates, and validity.

6 tasks

sequence task definitions
Task	Description	Expand details
Reverse-complement a CDS`dna.reverse_complement`	Given a CDS written in one direction, write the complementary DNA strand in its own 5′→3′ direction.
Reverse-complement a CDS DNA strands pair in opposite directions. The question supplies a CDS as a DNA string and asks for the sequence on the opposite strand. It does not ask where another gene starts or whether the opposite strand encodes a protein. Replace every base with its DNA partner, then reverse the order so the answer is written 5′→3′. Both steps matter: complementing alone leaves the answer in the wrong direction. Role in related tasks A reverse PCR primer must be written in the orientation that binds the opposite strand. CDS-internal primer selection uses this operation to turn the end of the desired amplicon into the reverse-primer sequence. Method reference: Primer3. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`. Procedure. Complement each of the 45 bases, then reverse the complete complemented string so the answer is written 5′→3′. Final answer: TCAGGAAGTGCGCCACCAACCTTTCAGTACGAAAATTGCTTTCAT Checking the answer. The answer must match the reverse complement base for base; whitespace between bases is ignored. Easy example · references supplied Prompt ## Task Return the reverse complement of the CDS, written 5′→3′. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Reference material ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: TCAGGAAGTGCGCCACCAACCTTTCAGTACGAAAATTGCTTTCAT` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Return the reverse complement of the CDS, written 5′→3′. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Reference material ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Return the reverse complement of the CDS, written 5′→3′. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: TCAGGAAGTGCGCCACCAACCTTTCAGTACGAAAATTGCTTTCAT` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Return the reverse complement of the CDS, written 5′→3′. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. An empty sequence is the token EMPTY_SEQUENCE. Formal fields and scoring Inputs `cds` DNA string Rules & boundaries Complement then reverse; no translation or ORF search. Scoring `exact_match` · complete parsed answer Reference blocks DNA reverse complement Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "dna.reverse_complement", "output_format": "scalar", "scorer": "exact_match", "family": "sequence", "domain": "dna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." }
Make the coding RNA`dna.transcribe`	Given coding-strand DNA, produce its coding RNA by replacing T with U while retaining the order of the bases.
Make the coding RNA During transcription, RNA polymerase copies the DNA template strand. The resulting RNA matches the opposite, coding strand except that RNA uses uracil (U) where DNA uses thymine (T). Here the supplied DNA is already that coding strand, written 5′→3′, with no introns to remove. Copy the DNA in the same order and replace every T with U. Keep the start codon and terminal stop codon in the RNA sequence. This creates the RNA input used by editing, pairing, and guide-matching tasks. Role in related tasks This is a representation step for following specified RNA edits through to a protein and matching RNA probes against a CDS collection. It supplies the coding RNA on which those operations act; it does not reconstruct a whole transcript. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`. Procedure. Copy all 45 coding-strand bases in order and replace T with U. The terminal TGA becomes UGA and remains in the RNA. Final answer: AUGAAAGCAAUUUUCGUACUGAAAGGUUGGUGGCGCACUUCCUGA Checking the answer. The complete RNA string must be correct and have the same length as the supplied CDS. Easy example · references supplied Prompt ## Task Return the coding RNA transcribed from this CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Retain the entire CDS, including the stop codon. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ## Answer - Return one continuous uppercase 5′→3′ RNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: AUGAAAGCAAUUUUCGUACUGAAAGGUUGGUGGCGCACUUCCUGA` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Return the coding RNA transcribed from this CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Retain the entire CDS, including the stop codon. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ## Answer - Return one continuous uppercase 5′→3′ RNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Return the coding RNA transcribed from this CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Retain the entire CDS, including the stop codon. ## Answer - Return one continuous uppercase 5′→3′ RNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: AUGAAAGCAAUUUUCGUACUGAAAGGUUGGUGGCGCACUUCCUGA` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Return the coding RNA transcribed from this CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Retain the entire CDS, including the stop codon. ## Answer - Return one continuous uppercase 5′→3′ RNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase 5′→3′ RNA sequence. An empty sequence is the token EMPTY_SEQUENCE. Formal fields and scoring Inputs `cds` DNA string Rules & boundaries T→U only; retain start, terminal stop, and length. Scoring `exact_match` · complete parsed answer Reference blocks Coding DNA to RNA Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "dna.transcribe", "output_format": "scalar", "scorer": "exact_match", "family": "sequence", "domain": "rna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: From specified RNA edits to protein From specified RNA edits to RNA fragments From a pairing annotation to eligible cuts Repair RNA pairs and verify the protein Balance an RNA structural constraint and codon preference Choose an RNA probe and a matched escape control Design and verify an RNA-editing-dependent product Build and audit a controlled synonymous panel From CDS collection to a minimal measurement panel
Translate DNA to protein`cds.translate`	Given a CDS and an identified genetic code, return the protein it encodes, ending at the first in-frame stop.
Translate DNA to protein The CDS is already spliced and oriented correctly. Its first base is the first base of the annotated initiation codon, so the model need not search for an open reading frame. The named genetic code determines how each three-base codon is interpreted. Read consecutive groups of three bases. Treat the annotated initiator according to NCBI code 11, and end translation at the first stop in that reading frame. Omit the stop and all downstream sequence from the protein answer. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`. Procedure. Read the 15 aligned codons from base 1: ATG starts M; the 13 subsequent sense codons complete MKAIFVLKGWWRTS; TGA ends translation and contributes no residue. Final answer: MKAIFVLKGWWRTS Checking the answer. The final amino-acid string must be exact. A correct prefix followed by extra residues is incorrect. Easy example · references supplied Prompt ## Task Translate this CDS using NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Use the annotated first codon as the initiator. Stop at the first in-frame stop and omit its symbol. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: MKAIFVLKGWWRTS` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Translate this CDS using NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Use the annotated first codon as the initiator. Stop at the first in-frame stop and omit its symbol. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Translate this CDS using NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Use the annotated first codon as the initiator. Stop at the first in-frame stop and omit its symbol. ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: MKAIFVLKGWWRTS` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Translate this CDS using NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Use the annotated first codon as the initiator. Stop at the first in-frame stop and omit its symbol. ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Role in composition tasks Protein editing and cloning-site cleanup: verifies that the final DNA encodes the requested protein. Tag assembly, cleavage, and mass calculation: supplies the protein sequence used by the downstream cleavage and mass calculations. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. An empty sequence is the token EMPTY_SEQUENCE. Formal fields and scoring Inputs `cds` DNA string Rules & boundaries NCBI 11; initiator→M; first in-frame stop terminates; no stop symbol. Scoring `exact_match` · complete parsed answer Reference blocks Annotated CDS initiation, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "cds.translate", "output_format": "scalar", "scorer": "exact_match", "family": "sequence", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Other composition tasks: From peptide motif to encoding DNA From CDS to defined protein measurements From CDS panel to a diagnostic peptide From DNA edits to changed peptide readouts Assemble coding modules and locate junction motifs Design and verify an RNA-editing-dependent product
Validate a complete CDS`cds.validate`	Given a candidate complete CDS, identify specific problems with its length, initiation codon, or stop codons.
Validate a complete CDS This is a check of a supplied sequence, not gene prediction. The candidate may have been damaged by an edit. A complete CDS must be long enough to contain an initiator and stop, have a length divisible by three, start with an allowed initiator, and end with a stop. Check each condition independently and report all violations. Also look for an earlier stop in the established reading frame. The formal labels below name these findings; they are not biological classes to infer from a sequence family. Related task: Protein-equivalence checking. Task Outline Inputs. `cds`: `GCT AAA GCA ATT TTC GTA CTG AAA GGT TGA TGG CGC ACT TCC GCA`. Procedure. Check the altered start, every aligned internal codon, and the last codon. The 45-base length is valid, but GCT is not an allowed initiator, codon 10 is an internal TGA, and the final GCA is not a stop. Final answer: ["internal_stop", "invalid_start", "missing_terminal_stop"] Checking the answer. The returned set of problem labels must be complete, without adding failures that the sequence does not have. Easy example · references supplied Prompt ## Task Validate this complete CDS under NCBI 11. Return all violation codes. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "GCT AAA GCA ATT TTC GTA CTG AAA GGT TGA TGG CGC ACT TCC GCA" } ``` ## Rules - Report `bad_length` if the DNA has fewer than 6 bases or its length is not divisible by 3. - Report `invalid_start` if the first codon is not an NCBI code 11 initiator. - Report `missing_terminal_stop` if the final codon is not a stop. - Report `internal_stop` if an earlier aligned codon is a stop. ## Reference material ### NCBI genetic code 11: allowed DNA starts The allowed annotated initiation codons are TTG, CTG, ATT, ATC, ATA, ATG, and GTG. ### NCBI genetic code 11: DNA stop codons The DNA stop codons are TAA, TAG, and TGA. All other unambiguous DNA codons are sense codons. ## Answer - Return a list of double-quoted violation labels, such as ["internal_stop", "missing_terminal_stop"]. - Include every applicable label once in ASCII order. The labels are bad_length, internal_stop, invalid_start, and missing_terminal_stop. A valid CDS gives []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ["internal_stop", "invalid_start", "missing_terminal_stop"]` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL with codon 1 changed to GCT, codon 10 to TGA, and terminal codon 15 to GCA, deliberately introducing validation faults. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Validate this complete CDS under NCBI 11. Return all violation codes. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Report `bad_length` if the DNA has fewer than 6 bases or its length is not divisible by 3. - Report `invalid_start` if the first codon is not an NCBI code 11 initiator. - Report `missing_terminal_stop` if the final codon is not a stop. - Report `internal_stop` if an earlier aligned codon is a stop. ## Reference material ### NCBI genetic code 11: allowed DNA starts The allowed annotated initiation codons are TTG, CTG, ATT, ATC, ATA, ATG, and GTG. ### NCBI genetic code 11: DNA stop codons The DNA stop codons are TAA, TAG, and TGA. All other unambiguous DNA codons are sense codons. ## Answer - Return a list of double-quoted violation labels, such as ["internal_stop", "missing_terminal_stop"]. - Include every applicable label once in ASCII order. The labels are bad_length, internal_stop, invalid_start, and missing_terminal_stop. A valid CDS gives []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Validate this complete CDS under NCBI 11. Return all violation codes. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "GCT AAA GCA ATT TTC GTA CTG AAA GGT TGA TGG CGC ACT TCC GCA" } ``` ## Rules - Report `bad_length` if the DNA has fewer than 6 bases or its length is not divisible by 3. - Report `invalid_start` if the first codon is not an NCBI code 11 initiator. - Report `missing_terminal_stop` if the final codon is not a stop. - Report `internal_stop` if an earlier aligned codon is a stop. ## Answer - Return a list of double-quoted violation labels, such as ["internal_stop", "missing_terminal_stop"]. - Include every applicable label once in ASCII order. The labels are bad_length, internal_stop, invalid_start, and missing_terminal_stop. A valid CDS gives []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ["internal_stop", "invalid_start", "missing_terminal_stop"]` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL with codon 1 changed to GCT, codon 10 to TGA, and terminal codon 15 to GCA, deliberately introducing validation faults. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Validate this complete CDS under NCBI 11. Return all violation codes. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Report `bad_length` if the DNA has fewer than 6 bases or its length is not divisible by 3. - Report `invalid_start` if the first codon is not an NCBI code 11 initiator. - Report `missing_terminal_stop` if the final codon is not a stop. - Report `internal_stop` if an earlier aligned codon is a stop. ## Answer - Return a list of double-quoted violation labels, such as ["internal_stop", "missing_terminal_stop"]. - Include every applicable label once in ASCII order. The labels are bad_length, internal_stop, invalid_start, and missing_terminal_stop. A valid CDS gives []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return a list of double-quoted violation labels, such as ["internal_stop", "missing_terminal_stop"]. Include every applicable label once in ASCII order. The labels are bad_length, internal_stop, invalid_start, and missing_terminal_stop. A valid CDS gives []. Formal fields and scoring Inputs `cds` DNA string Rules & boundaries Codes: bad_length if length<6 or not divisible by 3; invalid_start if first triplet is not an NCBI 11 initiator; missing_terminal_stop if final three bases are not a stop; internal_stop if an aligned stop precedes the final three bases. Return sorted codes; [] if valid. Scoring `exact_match` · complete parsed answer Reference blocks NCBI genetic code 11: allowed DNA starts, NCBI genetic code 11: DNA stop codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "cds.validate", "output_format": "scalar", "scorer": "exact_match", "family": "sequence", "domain": "dna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." }
Map a nucleotide to its codon`coords.nt_to_codon`	Given a nucleotide position in a CDS, identify its codon and its position within that codon.
Map a nucleotide to its codon The question identifies one DNA base by its position in the supplied CDS. Because translation begins at base 1, every base in the translated region belongs to a known group of three. Count codons from 1. Within the selected codon, use offsets 0, 1, and 2 for its first, second, and third base. Return the codon number, offset, and actual three DNA bases. This provides the bookkeeping needed to connect DNA mutations to protein residues. Role in related tasks A DNA-coordinate change needs to be located within a codon before its encoded consequence can be determined. Substitution-effect classification uses that connection to identify the affected residue and compare its original and edited codons. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `nt_position`: `29`. Procedure. Compute floor((29−1)/3)+1 = 10 and (29−1) mod 3 = 1. Read codon 10 from bases 28–30: TGG. Final answer: { "codon_index": 10, "offset": 1, "codon": "TGG" } Checking the answer. All three returned values must identify the same supplied nucleotide. Easy example Prompt ## Task For `nt_position`, report the containing codon, its index in the CDS, and the position of that nucleotide within the codon. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "nt_position": 29 } ``` ## Rules - Count codons from 1, starting with the first three bases of the CDS. - Within a codon, offsets 0, 1, and 2 mean its first, second, and third bases. - The supplied position lies before the terminal stop codon. ## Answer - `codon_index` is the 1-based aligned codon containing `nt_position`; `offset` is the base’s 0-based position within it. - `codon` is that complete codon as three uppercase DNA bases. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "codon_index": <positive integer>, "offset": <integer: 0, 1, or 2>, "codon": <three-base DNA string> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "codon_index": 10, "offset": 1, "codon": "TGG" }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task For `nt_position`, report the containing codon, its index in the CDS, and the position of that nucleotide within the codon. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "nt_position": {{nt_position}} } ``` ## Rules - Count codons from 1, starting with the first three bases of the CDS. - Within a codon, offsets 0, 1, and 2 mean its first, second, and third bases. - The supplied position lies before the terminal stop codon. ## Answer - `codon_index` is the 1-based aligned codon containing `nt_position`; `offset` is the base’s 0-based position within it. - `codon` is that complete codon as three uppercase DNA bases. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "codon_index": <positive integer>, "offset": <integer: 0, 1, or 2>, "codon": <three-base DNA string> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example Prompt ## Task For `nt_position`, report the containing codon, its index in the CDS, and the position of that nucleotide within the codon. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "nt_position": 29 } ``` ## Rules - Count codons from 1, starting with the first three bases of the CDS. - Within a codon, offsets 0, 1, and 2 mean its first, second, and third bases. - The supplied position lies before the terminal stop codon. ## Answer - `codon_index` is the 1-based aligned codon containing `nt_position`; `offset` is the base’s 0-based position within it. - `codon` is that complete codon as three uppercase DNA bases. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "codon_index": <positive integer>, "offset": <integer: 0, 1, or 2>, "codon": <three-base DNA string> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "codon_index": 10, "offset": 1, "codon": "TGG" }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task For `nt_position`, report the containing codon, its index in the CDS, and the position of that nucleotide within the codon. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "nt_position": {{nt_position}} } ``` ## Rules - Count codons from 1, starting with the first three bases of the CDS. - Within a codon, offsets 0, 1, and 2 mean its first, second, and third bases. - The supplied position lies before the terminal stop codon. ## Answer - `codon_index` is the 1-based aligned codon containing `nt_position`; `offset` is the base’s 0-based position within it. - `codon` is that complete codon as three uppercase DNA bases. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "codon_index": <positive integer>, "offset": <integer: 0, 1, or 2>, "codon": <three-base DNA string> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `codon_index` is the 1-based aligned codon containing `nt_position`; `offset` is the base’s 0-based position within it. `codon` is that complete codon as three uppercase DNA bases. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "codon_index": <positive integer>, "offset": <integer: 0, 1, or 2>, "codon": <three-base DNA string> } Formal fields and scoring Inputs `cds` DNA string `nt_position` integer within a sense codon Rules & boundaries Codon index is 1-based; offset=(position−1) mod 3. Position is before the terminal stop. Scoring `exact_match` · complete parsed answer Reference blocks None required. Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "coords.nt_to_codon", "output_format": "json", "scorer": "exact_match", "family": "sequence", "domain": "dna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." }
Map residues back to DNA`coords.aa_to_nt`	Given a range of protein residues, locate the DNA bases encoding that range in the original CDS.
Map residues back to DNA The question supplies a CDS and an interval of residue positions in its translated protein. It asks where that protein segment is encoded, not where a similar DNA string happens to occur. Each residue uses three consecutive bases. Include all three bases of the first and last residues in the requested range. Both ends of the returned DNA interval are included. The initiating methionine counts as residue 1. Role in related tasks Protein-engineering requests often name residues, while the construct must be changed in DNA. This mapping identifies the codons available for a requested amino-acid substitution or the exact interval removed in a protein deletion construct. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `aa_interval`: `[9, 12]`. Procedure. Residues 9–12 span four complete codons. Their first DNA base is 3×9−2 = 25 and their last is 3×12 = 36. Final answer: [25, 36] Checking the answer. The nucleotide interval must encode exactly the requested residue interval, with neither missing nor additional codons. Easy example · references supplied Prompt ## Task Find the DNA interval that encodes the requested protein interval `aa_interval`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "aa_interval": [9, 12] } ``` ## Rules - Count protein residues from 1, including the initiating residue. - Both endpoints of the requested protein interval are included. Return the DNA interval covering all their complete codons. ## Reference material ### Codon length A codon is a group of three consecutive nucleotides. ### Residue-to-codon correspondence Each amino-acid residue is encoded by one consecutive codon. The initiating residue is residue 1; a stop codon encodes no residue. ## Answer - Return [start, end], two integers giving the 1-based inclusive nucleotide interval in the original CDS. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: [25, 36]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find the DNA interval that encodes the requested protein interval `aa_interval`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "aa_interval": {{aa_interval}} } ``` ## Rules - Count protein residues from 1, including the initiating residue. - Both endpoints of the requested protein interval are included. Return the DNA interval covering all their complete codons. ## Reference material ### Codon length A codon is a group of three consecutive nucleotides. ### Residue-to-codon correspondence Each amino-acid residue is encoded by one consecutive codon. The initiating residue is residue 1; a stop codon encodes no residue. ## Answer - Return [start, end], two integers giving the 1-based inclusive nucleotide interval in the original CDS. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Find the DNA interval that encodes the requested protein interval `aa_interval`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "aa_interval": [9, 12] } ``` ## Rules - Count protein residues from 1, including the initiating residue. - Both endpoints of the requested protein interval are included. Return the DNA interval covering all their complete codons. ## Answer - Return [start, end], two integers giving the 1-based inclusive nucleotide interval in the original CDS. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: [25, 36]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find the DNA interval that encodes the requested protein interval `aa_interval`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "aa_interval": {{aa_interval}} } ``` ## Rules - Count protein residues from 1, including the initiating residue. - Both endpoints of the requested protein interval are included. Return the DNA interval covering all their complete codons. ## Answer - Return [start, end], two integers giving the 1-based inclusive nucleotide interval in the original CDS. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return [start, end], two integers giving the 1-based inclusive nucleotide interval in the original CDS. Formal fields and scoring Inputs `cds` DNA string `aa_interval` 1-based inclusive [a,b] within translated protein Rules & boundaries Residues a..b map to [3a−2,3b]; initiator is residue 1. Scoring `exact_match` · complete parsed answer Reference blocks Codon length, Residue-to-codon correspondence Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "coords.aa_to_nt", "output_format": "scalar", "scorer": "exact_match", "family": "sequence", "domain": "dna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: From peptide motif to encoding DNA Assemble coding modules and locate junction motifs

02 /

Sequence variation

Edits, ambiguity, and equivalence.

6 tasks

variation task definitions
Task	Description	Expand details
Apply a DNA edit script`variant.apply`	Given a CDS and a list of DNA replacements, insertions, or deletions, return the edited DNA sequence.
Apply a DNA edit script The question states exactly which edits to make. Every position refers to the original CDS, so an earlier insertion must not shift the coordinates used to interpret a later edit. No biological effect needs to be predicted in this task. A replacement names an inclusive interval and its new text; empty replacement text deletes that interval. An insertion names the boundary after an original base. Apply the nonoverlapping changes together and retain every unedited base. Role in related tasks Applying an edit script creates the candidate sequence that downstream checks must actually inspect. In DNA edits to peptide readouts, this step precedes translation and digestion, allowing the expected original and mutant peptide products to be compared. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `edits`: `[{"insert_after": 24, "sequence": "GGT"}, {"replace": [9, 9, "C"]}]`. Procedure. Insert GGT after original base 24 and change original base 9 to C. Apply both instructions against the original coordinates, so the insertion does not shift the replacement’s location. Final answer: ATGAAAGCCATTTTCGTACTGAAAGGTGGTTGGTGGCGCACTTCCTGA Checking the answer. The entire edited DNA string must be correct. Whether it still encodes a valid protein is checked by other tasks. Easy example Prompt ## Task Apply the supplied DNA edits and return the edited CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "edits": [ { "insert_after": 24, "sequence": "GGT" }, { "replace": [9, 9, "C"] } ] } ``` ## Rules - All positions refer to the original CDS. Do not renumber later edit coordinates after applying an earlier edit. - For an insertion after base k, put the new text between original bases k and k+1. Boundary 0 means before the first base. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCCATTTTCGTACTGAAAGGTGGTTGGTGGCGCACTTCCTGA` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Apply the supplied DNA edits and return the edited CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "edits": {{edits}} } ``` ## Rules - All positions refer to the original CDS. Do not renumber later edit coordinates after applying an earlier edit. - For an insertion after base k, put the new text between original bases k and k+1. Boundary 0 means before the first base. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example Prompt ## Task Apply the supplied DNA edits and return the edited CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "edits": [ { "insert_after": 24, "sequence": "GGT" }, { "replace": [9, 9, "C"] } ] } ``` ## Rules - All positions refer to the original CDS. Do not renumber later edit coordinates after applying an earlier edit. - For an insertion after base k, put the new text between original bases k and k+1. Boundary 0 means before the first base. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCCATTTTCGTACTGAAAGGTGGTTGGTGGCGCACTTCCTGA` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Apply the supplied DNA edits and return the edited CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "edits": {{edits}} } ``` ## Rules - All positions refer to the original CDS. Do not renumber later edit coordinates after applying an earlier edit. - For an insertion after base k, put the new text between original bases k and k+1. Boundary 0 means before the first base. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. An empty sequence is the token EMPTY_SEQUENCE. Formal fields and scoring Inputs `cds` DNA string `edits` nonoverlapping original-coordinate replacements/insertions Rules & boundaries Apply edits simultaneously using original coordinates; insertion boundaries follow the edit rules above. Scoring `exact_match` · complete parsed answer Reference blocks None required. Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "variant.apply", "output_format": "scalar", "scorer": "exact_match", "family": "variation", "domain": "dna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: From DNA edits to changed peptide readouts
Classify a substitution effect`variant.protein_effect`	Given an internal DNA substitution, determine whether it preserves the protein, changes residues, or introduces an early stop.
Classify a substitution effect The supplied change replaces DNA without changing its length and does not touch the initiator or terminal stop. Translate the original and edited CDSs using the same code and reading frame. An unchanged protein is synonymous. A change to one or more amino acids without an earlier stop is missense. A new earlier stop is stop_gained and takes precedence over other residue changes. Insertions, deletions, and start-codon changes are excluded so these three answers cover every instance. Practical context Synonymous, missense, and stop-gained are established sequence-consequence categories used by annotation tools such as Ensembl VEP. This task isolates their codon-comparison logic. The peptide-readout composition follows the sequence changes into expected digest products; neither task infers a variant’s functional impact. Method reference: Ensembl calculated variant consequences. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `replacement`: `[28, 30, "TGA"]`. Procedure. Replace Trp codon 10 TGG with TGA. Translation now ends after residue 9 instead of residue 14, so the earlier stop determines the class. Final answer: stop_gained Checking the answer. The selected category must follow from the actual translated products, rather than from the number of changed DNA bases. Easy example · references supplied Prompt ## Task Classify the supplied CDS substitution as `synonymous`, `missense`, or `stop_gained` under NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `replacement` is `[first_base, last_base, new_text]` in the original DNA coordinates. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "replacement": [28, 30, "TGA"] } ``` ## Rules - Translate the original CDS and the CDS after the replacement using NCBI genetic code 11. - If the replacement introduces an earlier in-frame stop, classify it as `stop_gained`, even if other residues also change. - Otherwise compare the complete protein strings: return `synonymous` if they are identical and `missense` if they differ. - The replacement has the same length as the removed interval and does not touch the first codon or terminal stop. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one lowercase token: synonymous, missense, or stop_gained. An earlier stop takes precedence over the other classes. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: stop_gained` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Classify the supplied CDS substitution as `synonymous`, `missense`, or `stop_gained` under NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `replacement` is `[first_base, last_base, new_text]` in the original DNA coordinates. ```json { "cds": "{{cds}}", "replacement": {{replacement}} } ``` ## Rules - Translate the original CDS and the CDS after the replacement using NCBI genetic code 11. - If the replacement introduces an earlier in-frame stop, classify it as `stop_gained`, even if other residues also change. - Otherwise compare the complete protein strings: return `synonymous` if they are identical and `missense` if they differ. - The replacement has the same length as the removed interval and does not touch the first codon or terminal stop. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one lowercase token: synonymous, missense, or stop_gained. An earlier stop takes precedence over the other classes. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Classify the supplied CDS substitution as `synonymous`, `missense`, or `stop_gained` under NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `replacement` is `[first_base, last_base, new_text]` in the original DNA coordinates. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "replacement": [28, 30, "TGA"] } ``` ## Rules - Translate the original CDS and the CDS after the replacement using NCBI genetic code 11. - If the replacement introduces an earlier in-frame stop, classify it as `stop_gained`, even if other residues also change. - Otherwise compare the complete protein strings: return `synonymous` if they are identical and `missense` if they differ. - The replacement has the same length as the removed interval and does not touch the first codon or terminal stop. ## Answer - Return one lowercase token: synonymous, missense, or stop_gained. An earlier stop takes precedence over the other classes. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: stop_gained` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Classify the supplied CDS substitution as `synonymous`, `missense`, or `stop_gained` under NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `replacement` is `[first_base, last_base, new_text]` in the original DNA coordinates. ```json { "cds": "{{cds}}", "replacement": {{replacement}} } ``` ## Rules - Translate the original CDS and the CDS after the replacement using NCBI genetic code 11. - If the replacement introduces an earlier in-frame stop, classify it as `stop_gained`, even if other residues also change. - Otherwise compare the complete protein strings: return `synonymous` if they are identical and `missense` if they differ. - The replacement has the same length as the removed interval and does not touch the first codon or terminal stop. ## Answer - Return one lowercase token: synonymous, missense, or stop_gained. An earlier stop takes precedence over the other classes. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one lowercase token: synonymous, missense, or stop_gained. An earlier stop takes precedence over the other classes. Formal fields and scoring Inputs `cds` valid complete DNA CDS `replacement` one equal-length replacement inside sense region; excludes first codon Rules & boundaries Compare original and edited translation. stop_gained takes precedence; otherwise identical=synonymous, different=missense. Restrict to substitutions so classes are exhaustive. Scoring `exact_match` · complete parsed answer Reference blocks Annotated CDS initiation, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "variant.protein_effect", "output_format": "scalar", "scorer": "exact_match", "family": "variation", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." }
Determine frame preservation`variant.frame_status`	Given several insertions and deletions, track whether the downstream reading frame is shifted or restored after each edit.
Determine frame preservation Adding or removing bases changes where subsequent codon boundaries fall. A net length change divisible by three preserves the downstream grouping; a remainder of one or two shifts it. An earlier shift can be undone by a later edit. Process edit locations in their original order and keep a running total of inserted bases minus deleted bases. After each edit, report that total modulo three: 0 means the original downstream frame, while 1 and 2 indicate the two shifted phases. This does not imply that the intervening protein is unchanged. Practical context This check helps review engineered insertions and deletions before interpreting their protein products. Even if a later edit restores the original frame, the intervening translation may have changed. Translation and peptide comparison examines the resulting product beyond this frame-status check. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `edits`: `[{"insert_after": 9, "sequence": "C"}, {"replace": [15, 15, ""]}]`. Procedure. The C insertion after original base 9 adds one base, giving phase 1. Deleting original base 15 removes one base and restores cumulative phase 0. Final answer: { "edits": [ { "edit_index": 1, "phase": 1 }, { "edit_index": 2, "phase": 0 } ], "final_phase": 0 } Checking the answer. Every intermediate phase and the final phase must match the cumulative length changes. Easy example Prompt ## Task Track the change in reading-frame phase after each insertion or deletion, and report the final phase. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "edits": [ { "insert_after": 9, "sequence": "C" }, { "replace": [15, 15, ""] } ] } ``` ## Rules - Process edits in order of their first affected original position. An insertion after base k is processed at position k+1. - If edits share a processing position, keep their original order in the input list. - After each edit, compute total inserted bases minus total deleted bases, reduced modulo 3 to a value of 0, 1, or 2. A later edit may restore phase 0. ## Answer - `edits` has one record per input edit, ordered by its first affected original base; an insertion after base k uses k+1. Break equal-coordinate ties by the original input index. - `edit_index` is the edit’s 1-based index in the input list. `phase` is cumulative (inserted bases minus deleted bases) modulo 3 immediately after that edit. - `final_phase` is the last phase, or 0 for an empty edit list. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "edits": [ { "edit_index": <positive integer>, "phase": <integer: 0, 1, or 2> } ], "final_phase": <integer: 0, 1, or 2> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "edits": [ { "edit_index": 1, "phase": 1 }, { "edit_index": 2, "phase": 0 } ], "final_phase": 0 }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Track the change in reading-frame phase after each insertion or deletion, and report the final phase. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "edits": {{edits}} } ``` ## Rules - Process edits in order of their first affected original position. An insertion after base k is processed at position k+1. - If edits share a processing position, keep their original order in the input list. - After each edit, compute total inserted bases minus total deleted bases, reduced modulo 3 to a value of 0, 1, or 2. A later edit may restore phase 0. ## Answer - `edits` has one record per input edit, ordered by its first affected original base; an insertion after base k uses k+1. Break equal-coordinate ties by the original input index. - `edit_index` is the edit’s 1-based index in the input list. `phase` is cumulative (inserted bases minus deleted bases) modulo 3 immediately after that edit. - `final_phase` is the last phase, or 0 for an empty edit list. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "edits": [ { "edit_index": <positive integer>, "phase": <integer: 0, 1, or 2> } ], "final_phase": <integer: 0, 1, or 2> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example Prompt ## Task Track the change in reading-frame phase after each insertion or deletion, and report the final phase. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "edits": [ { "insert_after": 9, "sequence": "C" }, { "replace": [15, 15, ""] } ] } ``` ## Rules - Process edits in order of their first affected original position. An insertion after base k is processed at position k+1. - If edits share a processing position, keep their original order in the input list. - After each edit, compute total inserted bases minus total deleted bases, reduced modulo 3 to a value of 0, 1, or 2. A later edit may restore phase 0. ## Answer - `edits` has one record per input edit, ordered by its first affected original base; an insertion after base k uses k+1. Break equal-coordinate ties by the original input index. - `edit_index` is the edit’s 1-based index in the input list. `phase` is cumulative (inserted bases minus deleted bases) modulo 3 immediately after that edit. - `final_phase` is the last phase, or 0 for an empty edit list. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "edits": [ { "edit_index": <positive integer>, "phase": <integer: 0, 1, or 2> } ], "final_phase": <integer: 0, 1, or 2> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "edits": [ { "edit_index": 1, "phase": 1 }, { "edit_index": 2, "phase": 0 } ], "final_phase": 0 }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Track the change in reading-frame phase after each insertion or deletion, and report the final phase. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "edits": {{edits}} } ``` ## Rules - Process edits in order of their first affected original position. An insertion after base k is processed at position k+1. - If edits share a processing position, keep their original order in the input list. - After each edit, compute total inserted bases minus total deleted bases, reduced modulo 3 to a value of 0, 1, or 2. A later edit may restore phase 0. ## Answer - `edits` has one record per input edit, ordered by its first affected original base; an insertion after base k uses k+1. Break equal-coordinate ties by the original input index. - `edit_index` is the edit’s 1-based index in the input list. `phase` is cumulative (inserted bases minus deleted bases) modulo 3 immediately after that edit. - `final_phase` is the last phase, or 0 for an empty edit list. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "edits": [ { "edit_index": <positive integer>, "phase": <integer: 0, 1, or 2> } ], "final_phase": <integer: 0, 1, or 2> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `edits` has one record per input edit, ordered by its first affected original base; an insertion after base k uses k+1. Break equal-coordinate ties by the original input index. `edit_index` is the edit’s 1-based index in the input list. `phase` is cumulative (inserted bases minus deleted bases) modulo 3 immediately after that edit. `final_phase` is the last phase, or 0 for an empty edit list. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "edits": [ { "edit_index": <positive integer>, "phase": <integer: 0, 1, or 2> } ], "final_phase": <integer: 0, 1, or 2> } Formal fields and scoring Inputs `cds` DNA string `edits` original-coordinate indels, all before terminal stop Rules & boundaries Process edit boundaries in increasing reference order. Phase is cumulative(inserted−deleted) mod 3; frame can be restored after a later edit. Equal-coordinate ties follow original input edit index; insertions after k sort at k+1. Scoring `exact_match` · complete parsed answer Reference blocks None required. Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "variant.frame_status", "output_format": "json", "scorer": "exact_match", "family": "variation", "domain": "dna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." }
Compare encoded proteins`cds.protein_equivalence`	Given two CDSs, determine whether their complete translated protein sequences are identical.
Compare encoded proteins Two different DNA sequences can encode the same protein because several codons can specify the same amino acid. DNA string equality therefore cannot answer this question. Translate both CDSs under NCBI code 11, including its initiator treatment and first-stop termination. Compare the resulting amino-acid strings, including their lengths. Return a yes/no answer; no judgement about expression or function is required. Practical context This is a correctness check for synonymous engineering: DNA changes are acceptable only if the encoded protein stays the same. Cloning-site removal with GC control needs this check so that manufacturing or assembly constraints do not silently alter the intended protein. Method reference: DNA Chisel examples. Task Outline Inputs. `cds_a`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `cds_b`: `ATG AAA GCC ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`. Procedure. Translate both full CDSs. GCA and GCC both encode Ala at residue 3, so both products are MKAIFVLKGWWRTS. Final answer: true Checking the answer. Both translated products must be identical for the answer to be true. Easy example · references supplied Prompt ## Task Do these two CDSs encode identical proteins under NCBI 11? ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_a": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "cds_b": "ATG AAA GCC ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Translate both complete CDSs under NCBI genetic code 11. For each, end at its first in-frame stop and omit the stop symbol. - Compare the complete resulting protein strings. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return the lowercase boolean true or false. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: true` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Compare native trpL with its synonymous GCA→GCC codon-3 variant. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Do these two CDSs encode identical proteins under NCBI 11? ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_a": "{{cds_a}}", "cds_b": "{{cds_b}}" } ``` ## Rules - Translate both complete CDSs under NCBI genetic code 11. For each, end at its first in-frame stop and omit the stop symbol. - Compare the complete resulting protein strings. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return the lowercase boolean true or false. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Do these two CDSs encode identical proteins under NCBI 11? ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_a": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "cds_b": "ATG AAA GCC ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Translate both complete CDSs under NCBI genetic code 11. For each, end at its first in-frame stop and omit the stop symbol. - Compare the complete resulting protein strings. ## Answer - Return the lowercase boolean true or false. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: true` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Compare native trpL with its synonymous GCA→GCC codon-3 variant. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Do these two CDSs encode identical proteins under NCBI 11? ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_a": "{{cds_a}}", "cds_b": "{{cds_b}}" } ``` ## Rules - Translate both complete CDSs under NCBI genetic code 11. For each, end at its first in-frame stop and omit the stop symbol. - Compare the complete resulting protein strings. ## Answer - Return the lowercase boolean true or false. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return the lowercase boolean true or false. Formal fields and scoring Inputs `cds_a` complete coding DNA `cds_b` complete coding DNA Rules & boundaries Translate each with NCBI 11 and first-stop termination; compare exact protein strings. Scoring `exact_match` · complete parsed answer Reference blocks Annotated CDS initiation, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "cds.protein_equivalence", "output_format": "scalar", "scorer": "exact_match", "family": "variation", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: Remove cloning sites while controlling GC Repair RNA pairs and verify the protein Build and audit a controlled synonymous panel
Resolve an ambiguous codon`ambiguity.codon_set`	Given an internal codon containing uncertain-base symbols, list every amino acid or stop it could encode.
Resolve an ambiguous codon An IUPAC symbol can stand for several possible DNA bases. For example, N means any of A, C, G, or T. The question identifies one ambiguous codon inside an otherwise defined CDS; its possible meanings are determined by expanding those symbols. Form every concrete codon allowed by the symbols, translate each using internal codon meanings, and collect the distinct results. Report each amino acid once, even if several codons produce it. Include STOP if any expansion is a stop. Role in related tasks An ambiguous triplet can represent uncertain base calls or a deliberately degenerate synthesis instruction. Degenerate-library enumeration uses the same expansion to establish exactly which amino acids and stops an oligonucleotide mixture could encode. Method reference: SwiftLib. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGN TGG CGC ACT TCC TGA`; `codon_index`: `10`. Procedure. Expand TGN to TGA, TGC, TGG, and TGT. Their internal products are STOP, C, W, and C; remove duplicate outcomes and sort the remaining strings. Final answer: ["C", "STOP", "W"] Checking the answer. The answer must contain every possible result and no result excluded by the ambiguity symbols. Easy example · references supplied Prompt ## Task List every amino acid or stop that the selected IUPAC codon could encode under NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGN TGG CGC ACT TCC TGA", "codon_index": 10 } ``` ## Rules - Inspect only the codon at `codon_index`. Consider every concrete codon allowed by its IUPAC symbols, expanding the three positions independently. - Use ordinary internal-codon meanings, not an initiation override. ## Reference material ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return a list of double-quoted one-letter amino-acid symbols; represent a stop by "STOP". - Include each possible internal-translation outcome once, sorted in ASCII order, including "STOP". After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ["C", "STOP", "W"]` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Replace native trpL codon 10 (TGG) with the ambiguous observation TGN; this denotes uncertainty, not a naturally occurring fifth base. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task List every amino acid or stop that the selected IUPAC codon could encode under NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "codon_index": {{codon_index}} } ``` ## Rules - Inspect only the codon at `codon_index`. Consider every concrete codon allowed by its IUPAC symbols, expanding the three positions independently. - Use ordinary internal-codon meanings, not an initiation override. ## Reference material ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return a list of double-quoted one-letter amino-acid symbols; represent a stop by "STOP". - Include each possible internal-translation outcome once, sorted in ASCII order, including "STOP". After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task List every amino acid or stop that the selected IUPAC codon could encode under NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGN TGG CGC ACT TCC TGA", "codon_index": 10 } ``` ## Rules - Inspect only the codon at `codon_index`. Consider every concrete codon allowed by its IUPAC symbols, expanding the three positions independently. - Use ordinary internal-codon meanings, not an initiation override. ## Answer - Return a list of double-quoted one-letter amino-acid symbols; represent a stop by "STOP". - Include each possible internal-translation outcome once, sorted in ASCII order, including "STOP". After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ["C", "STOP", "W"]` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Replace native trpL codon 10 (TGG) with the ambiguous observation TGN; this denotes uncertainty, not a naturally occurring fifth base. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task List every amino acid or stop that the selected IUPAC codon could encode under NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "codon_index": {{codon_index}} } ``` ## Rules - Inspect only the codon at `codon_index`. Consider every concrete codon allowed by its IUPAC symbols, expanding the three positions independently. - Use ordinary internal-codon meanings, not an initiation override. ## Answer - Return a list of double-quoted one-letter amino-acid symbols; represent a stop by "STOP". - Include each possible internal-translation outcome once, sorted in ASCII order, including "STOP". After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return a list of double-quoted one-letter amino-acid symbols; represent a stop by "STOP". Include each possible internal-translation outcome once, sorted in ASCII order, including "STOP". Formal fields and scoring Inputs `cds` CDS with exactly one ambiguous internal sense-position codon `codon_index` 1-based internal codon index Rules & boundaries Expand IUPAC symbols independently; use internal, not initiator, translation. Scoring `exact_match` · complete parsed answer Reference blocks IUPAC DNA base sets, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "ambiguity.codon_set", "output_format": "scalar", "scorer": "exact_match", "family": "variation", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." }
Assess an ambiguous stop`ambiguity.stop_certainty`	Given an ambiguous internal codon, decide whether a stop is impossible, possible, or unavoidable.
Assess an ambiguous stop The uncertain bases describe a set of possible concrete codons, not a probability distribution. The task asks what can be concluded about stopping from that set alone. Expand the selected codon. If none of its expansions is a stop, return never. If every expansion is a stop, return guaranteed. If both stop and non-stop expansions exist, return possible. This evaluates certainty without guessing which bases were actually present. Role in related tasks A mixed codon may produce a stop in only some synthesized molecules. This distinction matters in designing and auditing a mutagenesis library, where an apparently useful ambiguity code can also create unwanted truncated products. Method reference: SwiftLib. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGN TGG CGC ACT TCC TGA`; `codon_index`: `10`. Procedure. Of the four TGN expansions, TGA is a stop and the other three are sense codons. A stop is therefore possible but not guaranteed. Final answer: possible Checking the answer. The category must account for all allowed expansions, not just one representative codon. Easy example · references supplied Prompt ## Task Determine whether the selected ambiguous codon can never encode a stop, can sometimes encode a stop, or must always encode a stop under NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGN TGG CGC ACT TCC TGA", "codon_index": 10 } ``` ## Rules - Inspect only the codon at `codon_index` and consider all of its concrete IUPAC expansions. - Return `never` if none is a stop, `guaranteed` if all are stops, and `possible` otherwise. These are logical categories, not probabilities. ## Reference material ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### NCBI genetic code 11: DNA stop codons The DNA stop codons are TAA, TAG, and TGA. All other unambiguous DNA codons are sense codons. ## Answer - Return one lowercase token: never, possible, or guaranteed. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: possible` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Replace native trpL codon 10 (TGG) with the ambiguous observation TGN; this denotes uncertainty, not a naturally occurring fifth base. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Determine whether the selected ambiguous codon can never encode a stop, can sometimes encode a stop, or must always encode a stop under NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "codon_index": {{codon_index}} } ``` ## Rules - Inspect only the codon at `codon_index` and consider all of its concrete IUPAC expansions. - Return `never` if none is a stop, `guaranteed` if all are stops, and `possible` otherwise. These are logical categories, not probabilities. ## Reference material ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### NCBI genetic code 11: DNA stop codons The DNA stop codons are TAA, TAG, and TGA. All other unambiguous DNA codons are sense codons. ## Answer - Return one lowercase token: never, possible, or guaranteed. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Determine whether the selected ambiguous codon can never encode a stop, can sometimes encode a stop, or must always encode a stop under NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGN TGG CGC ACT TCC TGA", "codon_index": 10 } ``` ## Rules - Inspect only the codon at `codon_index` and consider all of its concrete IUPAC expansions. - Return `never` if none is a stop, `guaranteed` if all are stops, and `possible` otherwise. These are logical categories, not probabilities. ## Answer - Return one lowercase token: never, possible, or guaranteed. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: possible` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Replace native trpL codon 10 (TGG) with the ambiguous observation TGN; this denotes uncertainty, not a naturally occurring fifth base. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Determine whether the selected ambiguous codon can never encode a stop, can sometimes encode a stop, or must always encode a stop under NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "codon_index": {{codon_index}} } ``` ## Rules - Inspect only the codon at `codon_index` and consider all of its concrete IUPAC expansions. - Return `never` if none is a stop, `guaranteed` if all are stops, and `possible` otherwise. These are logical categories, not probabilities. ## Answer - Return one lowercase token: never, possible, or guaranteed. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one lowercase token: never, possible, or guaranteed. Formal fields and scoring Inputs `cds` CDS containing one ambiguous internal codon `codon_index` 1-based internal codon index Rules & boundaries never=no expanded codon is stop; guaranteed=all are stop; possible=some but not all. Scoring `exact_match` · complete parsed answer Reference blocks IUPAC DNA base sets, NCBI genetic code 11: DNA stop codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "ambiguity.stop_certainty", "output_format": "scalar", "scorer": "exact_match", "family": "variation", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: Design a codon library and verify its coverage

03 /

Sequence statistics

Separate, reproducible counts and numerical objectives.

6 tasks

measure task definitions
Task	Description	Expand details
Count sense codons`stats.codon_counts`	Given a CDS, count how often each DNA codon occurs in its translated region.
Count sense codons The CDS establishes the groups of three to count. This is different from counting every three-base substring: overlapping substrings between codon boundaries do not contribute. Count the initiator and all subsequent sense codons, but omit the terminal stop. Report the original DNA spellings, so synonymous codons such as GGT and GGC have separate counts even though both encode glycine. Role in related tasks Codon counts are the raw observations behind codon-usage comparisons and reference tables. Counts from a chosen reference gene set can support the preference weights used in CAI, while counts from a candidate CDS describe how its synonymous choices differ from that reference. Method reference: Codon Statistics Database. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`. Procedure. Read aligned codons through residue 14. AAA and TGG each occur twice; each other reported sense codon occurs once. Exclude the final TGA. Final answer: { "ATG": 1, "AAA": 2, "GCA": 1, "ATT": 1, "TTC": 1, "GTA": 1, "CTG": 1, "GGT": 1, "TGG": 2, "CGC": 1, "ACT": 1, "TCC": 1 } Checking the answer. Each observed codon’s count must be exact; the counts must sum to the number of sense codons. Easy example · references supplied Prompt ## Task Count how often each DNA sense codon occurs in the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Count aligned, nonoverlapping codons from the first base of the CDS. - Include the initiator and all subsequent sense codons. Exclude the terminal stop and omit codons whose count is zero. ## Reference material ### Codon length A codon is a group of three consecutive nucleotides. ## Answer - Map each observed uppercase DNA sense codon to its positive integer count among aligned codons. Include the initiator and omit the terminal stop and absent codons. - Each value counts occurrences of that codon, including repeats. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<DNA codon>": <positive integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "ATG": 1, "AAA": 2, "GCA": 1, "ATT": 1, "TTC": 1, "GTA": 1, "CTG": 1, "GGT": 1, "TGG": 2, "CGC": 1, "ACT": 1, "TCC": 1 }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Count how often each DNA sense codon occurs in the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Count aligned, nonoverlapping codons from the first base of the CDS. - Include the initiator and all subsequent sense codons. Exclude the terminal stop and omit codons whose count is zero. ## Reference material ### Codon length A codon is a group of three consecutive nucleotides. ## Answer - Map each observed uppercase DNA sense codon to its positive integer count among aligned codons. Include the initiator and omit the terminal stop and absent codons. - Each value counts occurrences of that codon, including repeats. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<DNA codon>": <positive integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Count how often each DNA sense codon occurs in the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Count aligned, nonoverlapping codons from the first base of the CDS. - Include the initiator and all subsequent sense codons. Exclude the terminal stop and omit codons whose count is zero. ## Answer - Map each observed uppercase DNA sense codon to its positive integer count among aligned codons. Include the initiator and omit the terminal stop and absent codons. - Each value counts occurrences of that codon, including repeats. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<DNA codon>": <positive integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "ATG": 1, "AAA": 2, "GCA": 1, "ATT": 1, "TTC": 1, "GTA": 1, "CTG": 1, "GGT": 1, "TGG": 2, "CGC": 1, "ACT": 1, "TCC": 1 }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Count how often each DNA sense codon occurs in the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Count aligned, nonoverlapping codons from the first base of the CDS. - Include the initiator and all subsequent sense codons. Exclude the terminal stop and omit codons whose count is zero. ## Answer - Map each observed uppercase DNA sense codon to its positive integer count among aligned codons. Include the initiator and omit the terminal stop and absent codons. - Each value counts occurrences of that codon, including repeats. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<DNA codon>": <positive integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` Map each observed uppercase DNA sense codon to its positive integer count among aligned codons. Include the initiator and omit the terminal stop and absent codons. Each value counts occurrences of that codon, including repeats. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "<DNA codon>": <positive integer count> } Formal fields and scoring Inputs `cds` valid complete DNA CDS Rules & boundaries Count all sense codons including initiator; exclude terminal stop; omit zero-count codons. Scoring `exact_match` · complete parsed answer Reference blocks Codon length Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "stats.codon_counts", "output_format": "json", "scorer": "exact_match", "family": "measure", "domain": "dna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." }
Calculate total GC`stats.gc_fraction`	Given a CDS, calculate the fraction of all its DNA bases that are G or C.
Calculate total GC GC content is a sequence-composition measurement. It counts bases across the complete supplied CDS, including its initiation and terminal stop codons. Add the number of G and C bases and divide by total sequence length. Return a fraction between 0 and 1, rounded as requested. This task does not ask whether that composition is biologically preferable. Practical context Overall GC is a quick composition check during synthetic-gene design. Manufacturing constraints often concern local extremes as well, so windowed GC checks and synonymous GC repair turn this simple measurement into a more useful construct-design check. Method reference: Twist gene-design resources. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`. Procedure. Count G and C across all 45 bases, including the initiator and terminal stop, then divide by 45 and round once to six decimal places. Final answer: 0.444444 Checking the answer. The number is checked against the calculated fraction at the stated precision. Easy example · references supplied Prompt ## Task Calculate the GC fraction of the complete CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Include every DNA base in the supplied CDS, including the first codon and terminal stop. ## Reference material ### GC fraction GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured. ## Answer - Return the fraction as a decimal number with 6 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 0.444444` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate the GC fraction of the complete CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Include every DNA base in the supplied CDS, including the first codon and terminal stop. ## Reference material ### GC fraction GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured. ## Answer - Return the fraction as a decimal number with 6 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Calculate the GC fraction of the complete CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Include every DNA base in the supplied CDS, including the first codon and terminal stop. ## Answer - Return the fraction as a decimal number with 6 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 0.444444` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate the GC fraction of the complete CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Include every DNA base in the supplied CDS, including the first codon and terminal stop. ## Answer - Return the fraction as a decimal number with 6 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return the fraction as a decimal number with 6 digits after the decimal point. Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. Formal fields and scoring Inputs `cds` DNA string Rules & boundaries Includes entire CDS, including terminal stop. Scoring `exact_match` · complete parsed answer Reference blocks GC fraction Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "stats.gc_fraction", "output_format": "scalar", "scorer": "exact_match", "family": "measure", "domain": "dna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: Build and audit a controlled synonymous panel
Calculate third-position GC`stats.gc3_fraction`	Given a CDS, calculate GC content using only the third base of each sense codon.
Calculate third-position GC Third codon positions are useful to examine separately because many synonymous choices differ there. The question asks for GC3, not the overall fraction of G and C in the CDS. Take the third base of every sense codon, including the initiator and excluding the terminal stop. Divide the number of those selected bases that are G or C by the number of selected bases. Other codon positions do not enter the calculation. Practical context GC3 is used in comparative codon-usage analysis to describe compositional bias at third codon positions, where many synonymous alternatives occur. Comparing it with other codon-bias statistics helps separate a sequence’s overall compositional tendency from more specific codon preferences. Method reference: Codon Statistics Database. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`. Procedure. Take the third base of each of the 14 sense codons. Seven are G or C, so GC3 is 7/14 = 0.500000. Final answer: 0.500000 Checking the answer. The answer must use the specified codon positions and denominator, with the requested rounding. Easy example · references supplied Prompt ## Task Calculate GC3 across the sense codons of the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Use the third base of every sense codon, including the initiator but excluding the terminal stop. - The denominator is the number of included codons, not the total DNA length. ## Reference material ### Codon length A codon is a group of three consecutive nucleotides. ### GC fraction GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured. ## Answer - Return the fraction as a decimal number with 6 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 0.500000` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate GC3 across the sense codons of the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Use the third base of every sense codon, including the initiator but excluding the terminal stop. - The denominator is the number of included codons, not the total DNA length. ## Reference material ### Codon length A codon is a group of three consecutive nucleotides. ### GC fraction GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured. ## Answer - Return the fraction as a decimal number with 6 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Calculate GC3 across the sense codons of the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Use the third base of every sense codon, including the initiator but excluding the terminal stop. - The denominator is the number of included codons, not the total DNA length. ## Answer - Return the fraction as a decimal number with 6 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 0.500000` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate GC3 across the sense codons of the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Use the third base of every sense codon, including the initiator but excluding the terminal stop. - The denominator is the number of included codons, not the total DNA length. ## Answer - Return the fraction as a decimal number with 6 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return the fraction as a decimal number with 6 digits after the decimal point. Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. Formal fields and scoring Inputs `cds` valid complete DNA CDS Rules & boundaries Include initiator, exclude terminal stop; denominator is number of sense codons. Scoring `exact_match` · complete parsed answer Reference blocks Codon length, GC fraction Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "stats.gc3_fraction", "output_format": "scalar", "scorer": "exact_match", "family": "measure", "domain": "dna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." }
Locate GC-window violations`stats.gc_windows`	Given a window length and an allowed GC range, identify local parts of a CDS whose composition falls outside that range.
Locate GC-window violations A whole-CDS average can hide a short GC-rich or GC-poor region. The question therefore gives a window length: the number of consecutive DNA bases to examine at a time. It also gives the lowest and highest acceptable GC fractions. Slide that window one base at a time across the CDS and calculate GC content separately for every full window. Return the starting positions of windows below the minimum or above the maximum. Values exactly on either bound are acceptable. Practical context A CDS can have acceptable overall GC but contain a locally extreme region that complicates synthesis or amplification. This measurement locates such regions; synonymous window repair changes them, and the cloning-and-GC composition combines that repair with assembly-site constraints. Method reference: Twist gene-design resources. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `window`: `12`; `gc_bounds`: `["1/4", "2/3"]`. Procedure. Slide a 12-base window by one base across the complete 45-base CDS. Count G/C in each of the 34 windows and report starts outside the inclusive 1/4–2/3 range. Final answer: [25] Checking the answer. The returned positions must identify all failing windows, including overlapping ones. Easy example · references supplied Prompt ## Task Find every full-length window whose GC fraction falls outside `gc_bounds`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. GC bounds are inclusive fractions. A string such as `"1/3"` denotes that exact rational value. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "window": 12, "gc_bounds": ["1/4", "2/3"] } ``` ## Rules - Move a window of length `window` along the full CDS one base at a time. Use only windows that fit completely within the sequence. - A fraction equal to either bound is allowed. Report starts only when the fraction is below the lower bound or above the upper bound. - Compare against the exact bounds before rounding. ## Reference material ### GC fraction GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured. ## Answer - Return a list of all qualifying 1-based integer start positions in increasing order, including overlapping windows. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: [25]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find every full-length window whose GC fraction falls outside `gc_bounds`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. GC bounds are inclusive fractions. A string such as `"1/3"` denotes that exact rational value. ```json { "cds": "{{cds}}", "window": {{window}}, "gc_bounds": {{gc_bounds}} } ``` ## Rules - Move a window of length `window` along the full CDS one base at a time. Use only windows that fit completely within the sequence. - A fraction equal to either bound is allowed. Report starts only when the fraction is below the lower bound or above the upper bound. - Compare against the exact bounds before rounding. ## Reference material ### GC fraction GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured. ## Answer - Return a list of all qualifying 1-based integer start positions in increasing order, including overlapping windows. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Find every full-length window whose GC fraction falls outside `gc_bounds`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. GC bounds are inclusive fractions. A string such as `"1/3"` denotes that exact rational value. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "window": 12, "gc_bounds": ["1/4", "2/3"] } ``` ## Rules - Move a window of length `window` along the full CDS one base at a time. Use only windows that fit completely within the sequence. - A fraction equal to either bound is allowed. Report starts only when the fraction is below the lower bound or above the upper bound. - Compare against the exact bounds before rounding. ## Answer - Return a list of all qualifying 1-based integer start positions in increasing order, including overlapping windows. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: [25]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find every full-length window whose GC fraction falls outside `gc_bounds`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. GC bounds are inclusive fractions. A string such as `"1/3"` denotes that exact rational value. ```json { "cds": "{{cds}}", "window": {{window}}, "gc_bounds": {{gc_bounds}} } ``` ## Rules - Move a window of length `window` along the full CDS one base at a time. Use only windows that fit completely within the sequence. - A fraction equal to either bound is allowed. Report starts only when the fraction is below the lower bound or above the upper bound. - Compare against the exact bounds before rounding. ## Answer - Return a list of all qualifying 1-based integer start positions in increasing order, including overlapping windows. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return a list of all qualifying 1-based integer start positions in increasing order, including overlapping windows. An empty result is []. Formal fields and scoring Inputs `cds` DNA string `window` integer width `gc_bounds` inclusive [minimum,maximum] fractions Rules & boundaries Every full window at stride 1; compare exact integer counts against bounds before rounding. Scoring `exact_match` · complete parsed answer Reference blocks GC fraction Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "stats.gc_windows", "output_format": "scalar", "scorer": "exact_match", "family": "measure", "domain": "dna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." }
Calculate codon adaptation index`stats.cai`	Given a CDS and codon-preference weights, calculate a codon adaptation index from those weights.
Calculate codon adaptation index The question supplies a numerical preference weight for each relevant codon. These weights define the reference against which the CDS is compared; the model is not expected to infer them from an organism name. Hosts differ in the availability of tRNAs that decode synonymous codons, which can affect translation efficiency. Highly expressed host genes provide an empirical reference for codon preferences. CAI measures resemblance to that reference, giving one guide for designing a gene for expression in that host. Broader methods such as LinearDesign combine CAI with RNA folding models to optimize codon usage and structural stability together. Look up the weight of each sense codon after the initiator and take their geometric mean. The initiator and terminal stop are excluded. The geometric mean combines preference across the sequence while retaining sensitivity to poorly weighted codons. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `weights`: supplied table shown in the prompt. Procedure. Read the six internal minigene codons and look up each supplied weight. Multiply those six values, take the sixth root, then round once; the initiator and terminal stop have no weight in this calculation. Final answer: 0.584804 Checking the answer. The calculated value must use the supplied weights and specified codon exclusions. Easy example · references supplied Prompt ## Task Calculate CAI using the supplied codon weights. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "weights": { "AAA": 1.0, "AAG": 0.3, "GCA": 0.2, "GCC": 0.3, "GCG": 1.0, "GCT": 0.5, "ATA": 0.2, "ATC": 0.3, "ATT": 1.0, "TTC": 1.0, "TTT": 0.3, "GTA": 0.2, "GTC": 0.3, "GTG": 1.0, "GTT": 0.5, "CTA": 0.2, "CTC": 0.3, "CTG": 1.0, "CTT": 0.5, "TTA": 0.6, "TTG": 0.7 } } ``` ## Rules - Calculate the geometric mean of the supplied codon weights, excluding the initiator and terminal stop. ## Reference material ### Codon length A codon is a group of three consecutive nucleotides. ### Codon adaptation index For n included codons with supplied positive weights w₁,…,wₙ, CAI = (w₁ × … × wₙ)^(1/n), the geometric mean of those weights. ## Answer - Return the CAI as a decimal number with 6 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 0.584804` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Supplied illustrative codon weights; these are not measured E. coli adaptation or expression estimates. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate CAI using the supplied codon weights. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "weights": {{weights}} } ``` ## Rules - Calculate the geometric mean of the supplied codon weights, excluding the initiator and terminal stop. ## Reference material ### Codon length A codon is a group of three consecutive nucleotides. ### Codon adaptation index For n included codons with supplied positive weights w₁,…,wₙ, CAI = (w₁ × … × wₙ)^(1/n), the geometric mean of those weights. ## Answer - Return the CAI as a decimal number with 6 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Calculate CAI using the supplied codon weights. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "weights": { "AAA": 1.0, "AAG": 0.3, "GCA": 0.2, "GCC": 0.3, "GCG": 1.0, "GCT": 0.5, "ATA": 0.2, "ATC": 0.3, "ATT": 1.0, "TTC": 1.0, "TTT": 0.3, "GTA": 0.2, "GTC": 0.3, "GTG": 1.0, "GTT": 0.5, "CTA": 0.2, "CTC": 0.3, "CTG": 1.0, "CTT": 0.5, "TTA": 0.6, "TTG": 0.7 } } ``` ## Rules - Calculate the geometric mean of the supplied codon weights, excluding the initiator and terminal stop. ## Answer - Return the CAI as a decimal number with 6 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 0.584804` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Supplied illustrative codon weights; these are not measured E. coli adaptation or expression estimates. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate CAI using the supplied codon weights. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "weights": {{weights}} } ``` ## Rules - Calculate the geometric mean of the supplied codon weights, excluding the initiator and terminal stop. ## Answer - Return the CAI as a decimal number with 6 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. This calculation is a prerequisite for synonymous codon-preference optimization, and is used in optimization with RNA constraints. Answer format and selection · scalar `output_format = scalar` Return the CAI as a decimal number with 6 digits after the decimal point. Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. Formal fields and scoring Inputs `cds` valid complete DNA CDS, at least 2 sense codons `weights` positive relative weights for every encountered codon Rules & boundaries Geometric mean; excludes first and stop codons. Custom weights remain present in hard prompts. Scoring `exact_match` · complete parsed answer Reference blocks Codon length, Codon adaptation index Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "stats.cai", "output_format": "scalar", "scorer": "exact_match", "family": "measure", "domain": "dna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." }
Find nucleotide patterns`dna.motif_positions`	Given a DNA pattern, locate every matching interval in the CDS on the requested strand or strands.
Find nucleotide patterns Restriction enzymes recognize particular DNA sequences: for example, EcoRI recognizes GAATTC. That is one concrete reason to locate a short pattern in a CDS. The pattern is supplied here and may include IUPAC symbols allowing several bases at a position. Check every possible starting position, including starts that cross codon boundaries and overlap other matches. If both strands are requested, also search for the reverse-complement pattern. Report all intervals on the coordinate axis of the original CDS, with the matching strand. Role in related tasks Locating restriction-enzyme recognition sites is a routine prerequisite for cloning-site cleanup. Motif removal uses these hits to eliminate unwanted sites, including sites crossing codon boundaries, while preserving the protein. Method reference: GoldenMutagenesis domestication. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `pattern`: `TGG`; `strand_mode`: `both`. Procedure. Search the forward DNA for TGG and for its reverse complement CCA. Report every match in original forward coordinates, including the two adjacent Trp codons. Final answer: [ { "start": 28, "end": 30, "strand": "+" }, { "start": 31, "end": 33, "strand": "+" } ] Checking the answer. Every matching interval and strand must be present. Identical hits from a palindromic pattern are reported only once. Easy example · references supplied Prompt ## Task Find every occurrence of the IUPAC DNA pattern on the strand or strands specified by `strand_mode`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "pattern": "TGG", "strand_mode": "both" } ``` ## Rules - Include overlapping matches. - With `strand_mode=forward`, search only the supplied coding strand. With `strand_mode=both`, also search for the reverse complement of the pattern. - Express every hit as an interval on the original forward CDS. A minus-strand hit still has its smaller forward coordinate first. - If the pattern equals its own reverse complement, report each matching interval once, with strand `+`. ## Reference material ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ## Answer - `start` and `end` are 1-based inclusive coordinates on the original forward CDS. `strand` is "+" or "-". - Include overlapping hits and sort by start, then end, then strand (+ before -). For a reverse-complement-palindromic pattern, report each interval once with strand "+". An empty result is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "strand": <"+" or "-"> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: [ { "start": 28, "end": 30, "strand": "+" }, { "start": 31, "end": 33, "strand": "+" } ]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find every occurrence of the IUPAC DNA pattern on the strand or strands specified by `strand_mode`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pattern": "{{pattern}}", "strand_mode": "{{strand_mode}}" } ``` ## Rules - Include overlapping matches. - With `strand_mode=forward`, search only the supplied coding strand. With `strand_mode=both`, also search for the reverse complement of the pattern. - Express every hit as an interval on the original forward CDS. A minus-strand hit still has its smaller forward coordinate first. - If the pattern equals its own reverse complement, report each matching interval once, with strand `+`. ## Reference material ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ## Answer - `start` and `end` are 1-based inclusive coordinates on the original forward CDS. `strand` is "+" or "-". - Include overlapping hits and sort by start, then end, then strand (+ before -). For a reverse-complement-palindromic pattern, report each interval once with strand "+". An empty result is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "strand": <"+" or "-"> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Find every occurrence of the IUPAC DNA pattern on the strand or strands specified by `strand_mode`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "pattern": "TGG", "strand_mode": "both" } ``` ## Rules - Include overlapping matches. - With `strand_mode=forward`, search only the supplied coding strand. With `strand_mode=both`, also search for the reverse complement of the pattern. - Express every hit as an interval on the original forward CDS. A minus-strand hit still has its smaller forward coordinate first. - If the pattern equals its own reverse complement, report each matching interval once, with strand `+`. ## Answer - `start` and `end` are 1-based inclusive coordinates on the original forward CDS. `strand` is "+" or "-". - Include overlapping hits and sort by start, then end, then strand (+ before -). For a reverse-complement-palindromic pattern, report each interval once with strand "+". An empty result is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "strand": <"+" or "-"> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: [ { "start": 28, "end": 30, "strand": "+" }, { "start": 31, "end": 33, "strand": "+" } ]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find every occurrence of the IUPAC DNA pattern on the strand or strands specified by `strand_mode`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pattern": "{{pattern}}", "strand_mode": "{{strand_mode}}" } ``` ## Rules - Include overlapping matches. - With `strand_mode=forward`, search only the supplied coding strand. With `strand_mode=both`, also search for the reverse complement of the pattern. - Express every hit as an interval on the original forward CDS. A minus-strand hit still has its smaller forward coordinate first. - If the pattern equals its own reverse complement, report each matching interval once, with strand `+`. ## Answer - `start` and `end` are 1-based inclusive coordinates on the original forward CDS. `strand` is "+" or "-". - Include overlapping hits and sort by start, then end, then strand (+ before -). For a reverse-complement-palindromic pattern, report each interval once with strand "+". An empty result is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "strand": <"+" or "-"> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `start` and `end` are 1-based inclusive coordinates on the original forward CDS. `strand` is "+" or "-". Include overlapping hits and sort by start, then end, then strand (+ before -). For a reverse-complement-palindromic pattern, report each interval once with strand "+". An empty result is []. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. [ { "start": <positive integer>, "end": <positive integer>, "strand": <"+" or "-"> } ] Formal fields and scoring Inputs `cds` DNA string `pattern` IUPAC DNA pattern `strand_mode` forward \| both Rules & boundaries Report overlapping matches. Minus matches are reverse-complement pattern hits on the forward coordinate axis; deduplicate identical palindromic hits as strand +. Scoring `exact_match` · complete parsed answer Reference blocks IUPAC DNA base sets, DNA reverse complement Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "dna.motif_positions", "output_format": "json", "scorer": "exact_match", "family": "measure", "domain": "dna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." }

04 /

Protein analysis

Patterns, properties, and peptide readouts derived from CDS.

9 tasks

protein task definitions
Task	Description	Expand details
Find a supplied peptide pattern`protein.motif_positions`	Translate the CDS and find every protein segment matching a supplied amino-acid pattern.
Find a supplied peptide pattern A supplied peptide pattern can describe a real enzyme-recognition site. For example, TEV protease cleavage uses ENLYFQG in this benchmark. More general patterns can name an exact residue, allow any residue, list alternatives, or exclude selected residues. Match the pattern at every protein position and return the residue intervals of complete matches. Include overlaps. This identifies a sequence pattern only; it does not require recognizing a protein family or deciding whether a matched site is functional. Role in related tasks An explicitly supplied peptide pattern can mark a protease site or an engineered sequence feature. Motif-to-DNA mapping locates its encoding codons for a later edit, and motif removal changes a permitted occurrence. The task does not infer a motif’s function from its sequence. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG AAT ACT TCC TGA`; `pattern`: `N{P}[ST]`. Procedure. Translate the R12N variant to MKAIFVLKGWWNTS. Test each three-residue window for N, then a residue other than P, then S or T. NTS at 12–14 is the match. Final answer: [[12, 14]] Checking the answer. The result must include exactly the matching residue intervals, numbered from the protein’s first residue. Easy example · references supplied Prompt ## Task Find every occurrence of the supplied peptide pattern in the protein encoded by the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG AAT ACT TCC TGA", "pattern": "N{P}[ST]" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. - Include overlapping full pattern matches. - Pattern syntax: a letter matches itself; `X` matches any residue; `[set]` matches a listed residue; `{set}` excludes listed residues. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return a list of [start, end] integer intervals in the translated protein, with 1-based inclusive endpoints. - Include every overlapping match once, sorted by start and then end. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: [[12, 14]]` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL with codon 12 CGC→AAT (R12N), creating the NTS sequence used to illustrate sequon matching; no actual glycosylation is asserted. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find every occurrence of the supplied peptide pattern in the protein encoded by the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pattern": "{{pattern}}" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. - Include overlapping full pattern matches. - Pattern syntax: a letter matches itself; `X` matches any residue; `[set]` matches a listed residue; `{set}` excludes listed residues. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return a list of [start, end] integer intervals in the translated protein, with 1-based inclusive endpoints. - Include every overlapping match once, sorted by start and then end. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Find every occurrence of the supplied peptide pattern in the protein encoded by the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG AAT ACT TCC TGA", "pattern": "N{P}[ST]" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. - Include overlapping full pattern matches. - Pattern syntax: a letter matches itself; `X` matches any residue; `[set]` matches a listed residue; `{set}` excludes listed residues. ## Answer - Return a list of [start, end] integer intervals in the translated protein, with 1-based inclusive endpoints. - Include every overlapping match once, sorted by start and then end. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: [[12, 14]]` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL with codon 12 CGC→AAT (R12N), creating the NTS sequence used to illustrate sequon matching; no actual glycosylation is asserted. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find every occurrence of the supplied peptide pattern in the protein encoded by the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pattern": "{{pattern}}" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. - Include overlapping full pattern matches. - Pattern syntax: a letter matches itself; `X` matches any residue; `[set]` matches a listed residue; `{set}` excludes listed residues. ## Answer - Return a list of [start, end] integer intervals in the translated protein, with 1-based inclusive endpoints. - Include every overlapping match once, sorted by start and then end. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return a list of [start, end] integer intervals in the translated protein, with 1-based inclusive endpoints. Include every overlapping match once, sorted by start and then end. An empty result is []. Formal fields and scoring Inputs `protein` derived protein string `pattern` residue tokens: letter, X=any, [set], or {excluded}; one residue per token Rules & boundaries Include overlaps; full pattern matches only. No functional inference from a match. Scoring `exact_match` · complete parsed answer Reference blocks Annotated CDS initiation, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "protein.motif_positions", "output_format": "scalar", "scorer": "exact_match", "family": "protein", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: From peptide motif to encoding DNA Assemble coding modules and locate junction motifs
Locate N-glycosylation sequons`protein.nglyco_sequons`	Translate the CDS and identify protein positions matching the conventional N-glycosylation sequon pattern.
Locate N-glycosylation sequons N-linked glycosylation attaches a sugar chain to an asparagine side chain in a protein. Its common acceptor pattern is asparagine, then any residue except proline, then serine or threonine: N–X–S/T with X≠P. A matching three-residue sequence is called a sequon. Find every occurrence of that three-residue pattern and report the position of its first residue. Do not decide whether the protein reaches a cellular compartment where glycosylation occurs or whether a particular site is actually modified. Practical context Sequon scanning identifies candidate sites to inspect when designing or reviewing a protein for N-glycosylation studies. It is a sequence screen: actual modification depends on the expression system and protein context. Residue-to-DNA mapping can locate the codons if a candidate site is selected for mutagenesis. Method reference: PROSITE N-glycosylation pattern. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG AAT ACT TCC TGA`. Procedure. Translate the R12N variant and scan for N–non-P–S/T. NTS begins at residue 12; report its start. This establishes a sequence match, not actual glycosylation. Final answer: [12] Checking the answer. All matching start positions must be correct, including overlapping matches. Easy example · references supplied Prompt ## Task List N-glycosylation sequon starts in the encoded protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG AAT ACT TCC TGA" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. - Include overlapping sequon matches. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### N-glycosylation sequon The conventional N-glycosylation sequon is N–X–S or N–X–T, where X is any amino-acid residue except P. ## Answer - Return a list of all matching 1-based integer start positions in increasing order, including overlapping matches. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: [12]` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL with codon 12 CGC→AAT (R12N), creating the NTS sequence used to illustrate sequon matching; no actual glycosylation is asserted. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task List N-glycosylation sequon starts in the encoded protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. - Include overlapping sequon matches. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### N-glycosylation sequon The conventional N-glycosylation sequon is N–X–S or N–X–T, where X is any amino-acid residue except P. ## Answer - Return a list of all matching 1-based integer start positions in increasing order, including overlapping matches. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task List N-glycosylation sequon starts in the encoded protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG AAT ACT TCC TGA" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. - Include overlapping sequon matches. ## Answer - Return a list of all matching 1-based integer start positions in increasing order, including overlapping matches. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: [12]` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL with codon 12 CGC→AAT (R12N), creating the NTS sequence used to illustrate sequon matching; no actual glycosylation is asserted. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task List N-glycosylation sequon starts in the encoded protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. - Include overlapping sequon matches. ## Answer - Return a list of all matching 1-based integer start positions in increasing order, including overlapping matches. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return a list of all matching 1-based integer start positions in increasing order, including overlapping matches. An empty result is []. Formal fields and scoring Inputs `protein` derived protein string Rules & boundaries X≠P; include overlaps. Pattern membership only. Scoring `exact_match` · complete parsed answer Reference blocks Annotated CDS initiation, NCBI genetic code 11: internal DNA codons, N-glycosylation sequon Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "protein.nglyco_sequons", "output_format": "scalar", "scorer": "exact_match", "family": "protein", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." }
Count amino acids`protein.residue_counts`	Translate a CDS and count the number of each amino acid in its protein product.
Count amino acids This measures protein composition rather than DNA composition or codon usage. Different synonymous codons contribute to the same amino-acid count after translation. Translate the CDS, then count each residue in the resulting protein. Include the initiating methionine and do not count the stop. Return counts for residues that occur; the order of residues in the protein is not part of this answer. Role in related tasks Composition counts supply inputs to mass, charge, and isoelectric-point calculations. The protein-measurements composition connects a CDS to these descriptors, which can support product identification and purification planning. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`. Procedure. Translate to MKAIFVLKGWWRTS and increment one count per residue. K and W each contribute two occurrences; every other observed residue contributes one. Final answer: { "M": 1, "K": 2, "A": 1, "I": 1, "F": 1, "V": 1, "L": 1, "G": 1, "W": 2, "R": 1, "T": 1, "S": 1 } Checking the answer. The residue counts must be exact and sum to the translated protein’s length. Easy example · references supplied Prompt ## Task Return the encoded protein’s amino-acid counts. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Map each observed uppercase one-letter amino-acid symbol to its positive integer count in the complete translated protein. Omit absent residues and stops. An empty protein gives {}. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<one-letter amino-acid symbol>": <positive integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "M": 1, "K": 2, "A": 1, "I": 1, "F": 1, "V": 1, "L": 1, "G": 1, "W": 2, "R": 1, "T": 1, "S": 1 }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Return the encoded protein’s amino-acid counts. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Map each observed uppercase one-letter amino-acid symbol to its positive integer count in the complete translated protein. Omit absent residues and stops. An empty protein gives {}. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<one-letter amino-acid symbol>": <positive integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Return the encoded protein’s amino-acid counts. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. ## Answer - Map each observed uppercase one-letter amino-acid symbol to its positive integer count in the complete translated protein. Omit absent residues and stops. An empty protein gives {}. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<one-letter amino-acid symbol>": <positive integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "M": 1, "K": 2, "A": 1, "I": 1, "F": 1, "V": 1, "L": 1, "G": 1, "W": 2, "R": 1, "T": 1, "S": 1 }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Return the encoded protein’s amino-acid counts. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. ## Answer - Map each observed uppercase one-letter amino-acid symbol to its positive integer count in the complete translated protein. Omit absent residues and stops. An empty protein gives {}. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<one-letter amino-acid symbol>": <positive integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` Map each observed uppercase one-letter amino-acid symbol to its positive integer count in the complete translated protein. Omit absent residues and stops. An empty protein gives {}. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "<one-letter amino-acid symbol>": <positive integer count> } Formal fields and scoring Inputs `protein` derived protein string Rules & boundaries Count residues, not codons; omit absent residues and stops. Scoring `exact_match` · complete parsed answer Reference blocks Annotated CDS initiation, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "protein.residue_counts", "output_format": "json", "scorer": "exact_match", "family": "protein", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: From CDS to defined protein measurements
Calculate peptide mass`protein.monoisotopic_mass`	Translate a CDS and calculate the neutral monoisotopic mass of its unmodified protein product.
Calculate peptide mass Mass calculations depend on a chemical convention. This task specifies an unmodified linear protein with free ends, without removing the initiator or forming disulfide bonds. Monoisotopic mass uses the masses of the specified isotopes rather than an average over isotopes. A chain of n amino acids has n−1 peptide bonds; its chemical composition differs from n free amino acids by n−1 water molecules. The tabulated residue masses already subtract one water per residue. Sum those residue masses and add one water to recover the complete chain with free ends. Each separate digestion product gets its own water contribution. Practical context Theoretical peptide masses help interpret mass-spectrometry measurements when checking a protein’s identity or processing. Tag assembly, cleavage, and mass calculation applies this to the actual cleavage products, where a retained tag or cleavage scar changes the expected mass. Method reference: ExPASy PeptideMass documentation. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`. Procedure. Translate the 14-residue leader peptide. Add the supplied monoisotopic residue masses and one water molecule for the complete unmodified linear peptide, then round to four decimal places. Final answer: 1721.9440 Checking the answer. The declared residue-mass sum plus one water must round to the reported four-decimal mass. Easy example · references supplied Prompt ## Task Calculate the neutral monoisotopic mass of the complete protein encoded by the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - Use an unmodified linear protein with both termini free. Keep the initiating residue and assume no disulfide bonds. - Use the mass constants below, including the complete peptide's free termini. ## Mass constants Use these isotope masses in Da as exact decimal constants: ```text 12C = 12.00000000000 1H = 1.00782503223 14N = 14.00307400443 16O = 15.99491461957 32S = 31.9720711744 ``` For a neutral, unmodified linear peptide, sum the neutral free-amino-acid molecular formulas and subtract one H2O for each peptide bond. Equivalently, sum the incorporated residue masses and add one H2O for the complete peptide. Do not round intermediate values. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Monoisotopic residue masses Use these incorporated residue masses in Da as exact decimal constants. Each is the corresponding neutral free amino acid minus one H2O. ```text A=71.03711378515 C=103.00918495955 D=115.02694302429 E=129.04259308875 F=147.06841391407 G=57.02146372069 H=137.05891185847 I=113.08406397853 K=128.09496301519 L=113.08406397853 M=131.04048508847 N=114.04292744138 P=97.05276384961 Q=128.05857750584 R=156.10111102405 S=87.03202840472 T=101.04767846918 V=99.06841391407 W=186.07931295073 Y=163.06332853364 ``` Add one H2O = 18.01056468403 Da for the complete free linear peptide. ## Answer - Return the mass in Da as a decimal number with 4 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 1721.9440` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate the neutral monoisotopic mass of the complete protein encoded by the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - Use an unmodified linear protein with both termini free. Keep the initiating residue and assume no disulfide bonds. - Use the mass constants below, including the complete peptide's free termini. ## Mass constants Use these isotope masses in Da as exact decimal constants: ```text 12C = 12.00000000000 1H = 1.00782503223 14N = 14.00307400443 16O = 15.99491461957 32S = 31.9720711744 ``` For a neutral, unmodified linear peptide, sum the neutral free-amino-acid molecular formulas and subtract one H2O for each peptide bond. Equivalently, sum the incorporated residue masses and add one H2O for the complete peptide. Do not round intermediate values. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Monoisotopic residue masses Use these incorporated residue masses in Da as exact decimal constants. Each is the corresponding neutral free amino acid minus one H2O. ```text A=71.03711378515 C=103.00918495955 D=115.02694302429 E=129.04259308875 F=147.06841391407 G=57.02146372069 H=137.05891185847 I=113.08406397853 K=128.09496301519 L=113.08406397853 M=131.04048508847 N=114.04292744138 P=97.05276384961 Q=128.05857750584 R=156.10111102405 S=87.03202840472 T=101.04767846918 V=99.06841391407 W=186.07931295073 Y=163.06332853364 ``` Add one H2O = 18.01056468403 Da for the complete free linear peptide. ## Answer - Return the mass in Da as a decimal number with 4 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Calculate the neutral monoisotopic mass of the complete protein encoded by the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - Use an unmodified linear protein with both termini free. Keep the initiating residue and assume no disulfide bonds. - Use the mass constants below, including the complete peptide's free termini. ## Mass constants Use these isotope masses in Da as exact decimal constants: ```text 12C = 12.00000000000 1H = 1.00782503223 14N = 14.00307400443 16O = 15.99491461957 32S = 31.9720711744 ``` For a neutral, unmodified linear peptide, sum the neutral free-amino-acid molecular formulas and subtract one H2O for each peptide bond. Equivalently, sum the incorporated residue masses and add one H2O for the complete peptide. Do not round intermediate values. ## Answer - Return the mass in Da as a decimal number with 4 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 1721.9440` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate the neutral monoisotopic mass of the complete protein encoded by the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - Use an unmodified linear protein with both termini free. Keep the initiating residue and assume no disulfide bonds. - Use the mass constants below, including the complete peptide's free termini. ## Mass constants Use these isotope masses in Da as exact decimal constants: ```text 12C = 12.00000000000 1H = 1.00782503223 14N = 14.00307400443 16O = 15.99491461957 32S = 31.9720711744 ``` For a neutral, unmodified linear peptide, sum the neutral free-amino-acid molecular formulas and subtract one H2O for each peptide bond. Equivalently, sum the incorporated residue masses and add one H2O for the complete peptide. Do not round intermediate values. ## Answer - Return the mass in Da as a decimal number with 4 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return the mass in Da as a decimal number with 4 digits after the decimal point. Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. Formal fields and scoring Inputs `protein` derived protein string Rules & boundaries Unmodified linear protein, free termini, no initiator removal, no disulfides; use the exact supplied isotope constants, or the derived incorporated residue sum plus one H2O. Round only the final peptide mass to 4 decimals. Scoring `exact_match` · complete parsed answer Reference blocks Annotated CDS initiation, NCBI genetic code 11: internal DNA codons, Monoisotopic isotope masses, Monoisotopic residue masses Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "protein.monoisotopic_mass", "output_format": "scalar", "scorer": "exact_match", "family": "protein", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: From CDS to defined protein measurements Build a tagged construct, cleave it, measure products
Calculate mean hydropathy`protein.mean_hydropathy`	Translate a CDS and average the encoded residues’ scores on the Kyte–Doolittle hydropathy scale.
Calculate mean hydropathy Amino-acid side chains differ in how favorably they interact with water. The Kyte–Doolittle hydropathy scale assigns experimentally informed scores to that tendency: positive values are more hydrophobic, while negative values are more hydrophilic. These are a published scale, not arbitrary scores invented for this task. Look up the score for each residue and calculate the ordinary arithmetic mean across the whole protein. This is a summary measurement; it is not a request to classify the protein as a transporter or predict its location. Practical context Mean hydropathy is a standard descriptive property reported by tools such as ProtParam. It summarizes overall residue composition; hydropathy analyses that seek local hydrophobic segments use positional windows instead. Here it belongs in a protein-characterization report, rather than serving as a standalone membrane-localization prediction. Method reference: ExPASy ProtParam documentation. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`. Procedure. Translate the 14-residue leader peptide, add its 14 Kyte–Doolittle scores, and divide by 14. Count both Lys and both Trp residues in the mean. Final answer: 0.2143 Checking the answer. The answer must use the named scale and average all translated residues with equal weight. Easy example · references supplied Prompt ## Task Calculate the whole-protein mean Kyte–Doolittle hydropathy encoded by this CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Kyte–Doolittle hydropathy scores Use these Kyte–Doolittle residue scores. Whole-protein mean hydropathy is their arithmetic mean, counting every residue occurrence. ```text A=1.8 R=-4.5 N=-3.5 D=-3.5 C=2.5 Q=-3.5 E=-3.5 G=-0.4 H=-3.2 I=4.5 L=3.8 K=-3.9 M=1.9 F=2.8 P=-1.6 S=-0.8 T=-0.7 W=-0.9 Y=-1.3 V=4.2 ``` ## Answer - Return the mean hydropathy as a decimal number with 4 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 0.2143` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate the whole-protein mean Kyte–Doolittle hydropathy encoded by this CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Kyte–Doolittle hydropathy scores Use these Kyte–Doolittle residue scores. Whole-protein mean hydropathy is their arithmetic mean, counting every residue occurrence. ```text A=1.8 R=-4.5 N=-3.5 D=-3.5 C=2.5 Q=-3.5 E=-3.5 G=-0.4 H=-3.2 I=4.5 L=3.8 K=-3.9 M=1.9 F=2.8 P=-1.6 S=-0.8 T=-0.7 W=-0.9 Y=-1.3 V=4.2 ``` ## Answer - Return the mean hydropathy as a decimal number with 4 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Calculate the whole-protein mean Kyte–Doolittle hydropathy encoded by this CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. ## Answer - Return the mean hydropathy as a decimal number with 4 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 0.2143` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate the whole-protein mean Kyte–Doolittle hydropathy encoded by this CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. ## Answer - Return the mean hydropathy as a decimal number with 4 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return the mean hydropathy as a decimal number with 4 digits after the decimal point. Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. Formal fields and scoring Inputs `protein` derived protein string Rules & boundaries Arithmetic mean across all residues. Round to 4 decimals. Scoring `exact_match` · complete parsed answer Reference blocks Annotated CDS initiation, NCBI genetic code 11: internal DNA codons, Kyte–Doolittle hydropathy scores Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "protein.mean_hydropathy", "output_format": "scalar", "scorer": "exact_match", "family": "protein", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: From CDS to defined protein measurements
Digest with trypsin rules`protein.trypsin_digest`	Translate the CDS and list the peptide fragments produced by a specified complete trypsin-cleavage model.
Digest with trypsin rules Trypsin is a real protein-digesting enzyme used to generate peptides for mass spectrometry. Its specificity favors cutting after lysine (K) or arginine (R). The benchmark uses the familiar simplified rule: cut after K/R unless the following residue is proline (P). Actual cleavage efficiencies and exceptions are not calculated. Apply the rule at every eligible bond, allowing no missed cuts, and retain fragments at both ends. Return fragments in their original order, together with their inclusive residue intervals in the original protein. Practical context In-silico digestion predicts peptide sequences to compare with proteomics measurements. Diagnostic-peptide selection looks for products that distinguish related proteins, while variant-to-peptide comparison shows how a sequence change alters the expected digest. Method reference: ExPASy PeptideMass documentation. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`. Procedure. Translate to MKAIFVLKGWWRTS. Cut after K2, K8, and R12. This gives MK, AIFVLK, GWWR, and TS; retain the nonempty final TS fragment. Final answer: [ { "start": 1, "end": 2, "peptide": "MK" }, { "start": 3, "end": 8, "peptide": "AIFVLK" }, { "start": 9, "end": 12, "peptide": "GWWR" }, { "start": 13, "end": 14, "peptide": "TS" } ] Checking the answer. Both peptide sequences and their original residue intervals must be correct. Easy example · references supplied Prompt ## Task List the fragments produced by a complete trypsin digest of the encoded protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. - Use complete trypsin digestion with zero missed cleavage sites: cut every site that meets the trypsin specificity rule. - Keep the nonempty fragments at both ends, even if the protein has no eligible cut sites. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Trypsin cleavage specificity Trypsin cleaves after K or R, except when the next residue is P. The K or R remains on the upstream peptide. ## Answer - `start` and `end` are 1-based inclusive residue coordinates in the complete translated protein; `peptide` is the fragment’s continuous uppercase sequence. - List every nonempty fragment in N-to-C order, retaining identical sequences from different intervals. An empty protein gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "peptide": <amino-acid string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: [ { "start": 1, "end": 2, "peptide": "MK" }, { "start": 3, "end": 8, "peptide": "AIFVLK" }, { "start": 9, "end": 12, "peptide": "GWWR" }, { "start": 13, "end": 14, "peptide": "TS" } ]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task List the fragments produced by a complete trypsin digest of the encoded protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. - Use complete trypsin digestion with zero missed cleavage sites: cut every site that meets the trypsin specificity rule. - Keep the nonempty fragments at both ends, even if the protein has no eligible cut sites. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Trypsin cleavage specificity Trypsin cleaves after K or R, except when the next residue is P. The K or R remains on the upstream peptide. ## Answer - `start` and `end` are 1-based inclusive residue coordinates in the complete translated protein; `peptide` is the fragment’s continuous uppercase sequence. - List every nonempty fragment in N-to-C order, retaining identical sequences from different intervals. An empty protein gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "peptide": <amino-acid string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task List the fragments produced by a complete trypsin digest of the encoded protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. - Use complete trypsin digestion with zero missed cleavage sites: cut every site that meets the trypsin specificity rule. - Keep the nonempty fragments at both ends, even if the protein has no eligible cut sites. ## Answer - `start` and `end` are 1-based inclusive residue coordinates in the complete translated protein; `peptide` is the fragment’s continuous uppercase sequence. - List every nonempty fragment in N-to-C order, retaining identical sequences from different intervals. An empty protein gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "peptide": <amino-acid string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: [ { "start": 1, "end": 2, "peptide": "MK" }, { "start": 3, "end": 8, "peptide": "AIFVLK" }, { "start": 9, "end": 12, "peptide": "GWWR" }, { "start": 13, "end": 14, "peptide": "TS" } ]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task List the fragments produced by a complete trypsin digest of the encoded protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}" } ``` ## Rules - Translate the CDS with NCBI genetic code 11, stopping at the first in-frame stop. - Use complete trypsin digestion with zero missed cleavage sites: cut every site that meets the trypsin specificity rule. - Keep the nonempty fragments at both ends, even if the protein has no eligible cut sites. ## Answer - `start` and `end` are 1-based inclusive residue coordinates in the complete translated protein; `peptide` is the fragment’s continuous uppercase sequence. - List every nonempty fragment in N-to-C order, retaining identical sequences from different intervals. An empty protein gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "peptide": <amino-acid string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `start` and `end` are 1-based inclusive residue coordinates in the complete translated protein; `peptide` is the fragment’s continuous uppercase sequence. List every nonempty fragment in N-to-C order, retaining identical sequences from different intervals. An empty protein gives []. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. [ { "start": <positive integer>, "end": <positive integer>, "peptide": <amino-acid string> } ] Formal fields and scoring Inputs `protein` derived protein string Rules & boundaries Named benchmark rule trypsin_KR_notP_v1; complete digestion; zero missed cleavages. Scoring `exact_match` · complete parsed answer Reference blocks Annotated CDS initiation, NCBI genetic code 11: internal DNA codons, Trypsin cleavage specificity Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "protein.trypsin_digest", "output_format": "json", "scorer": "exact_match", "family": "protein", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: From CDS panel to a diagnostic peptide From DNA edits to changed peptide readouts
Select a distinguishing peptide`protein.diagnostic_peptide`	From a target CDS and comparison CDSs, choose a digest peptide that occurs only in the target’s protein.
Select a distinguishing peptide You are given a target CDS encoding the protein you want to identify, and comparison CDSs encoding other proteins you want to distinguish it from. The comparison set is called the background in the input fields. Translate and digest each sequence using the same trypsin rule. An identifying peptide must occur in the target digest and in none of the comparison digests. Keep target peptides whose lengths lie within the supplied minimum and maximum. Select the shortest eligible peptide, using alphabetical order to break a tie. Return NONE if none qualifies. Uniqueness applies only to the supplied comparison set, not an entire proteome. Practical context A peptide unique to one candidate protein is a starting point for choosing an identifying mass-spectrometry readout. The CDS-panel composition derives such candidates from actual digests. Subsequent assay design must also consider whether the peptide can be detected reliably, not just whether its sequence is unique. Method reference: ExPASy PeptideMass documentation. Task Outline Inputs. `target_cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `background_cds_panel`: `["ATG AAA GCA GTT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA"]`; `length_bounds`: `[4, 10]`. Procedure. The target digest is MK, AIFVLK, GWWR, TS. The I4V comparison changes only AIFVLK to AVFVLK. Remove shared peptides and apply the length limits; AIFVLK is the qualifying target-only sequence. Final answer: AIFVLK Checking the answer. The selected peptide must satisfy panel uniqueness and length limits, and obey the stated selection order. Detectability in a mass spectrometer is not scored. Easy example · references supplied Prompt ## Task Choose a peptide that distinguishes the target protein from every protein in the comparison panel. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "target_cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "background_cds_panel": ["ATG AAA GCA GTT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA"], "length_bounds": [4, 10] } ``` ## Rules - Translate the target and all comparison CDSs under NCBI genetic code 11, stopping each at its first in-frame stop. - Digest every protein completely with trypsin, with zero missed cleavage sites. - A peptide qualifies only if its sequence occurs in the target digest and in none of the comparison digests. Ignore how many copies occur. - The peptide length must be between both inclusive `length_bounds`. ## Selection From eligible target-only peptide sequences, choose the shortest, then the ASCII-lexicographically smallest sequence. Multiple copies of one sequence do not create separate choices. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Trypsin cleavage specificity Trypsin cleaves after K or R, except when the next residue is P. The K or R remains on the upstream peptide. ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. If no target-only peptide meets the length limits, return: ```text NONE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: AIFVLK` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL is the target. The comparison is a designed I4V variant (codon 4 ATT→GTT), not a second natural accession. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Choose a peptide that distinguishes the target protein from every protein in the comparison panel. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "target_cds": "{{target_cds}}", "background_cds_panel": {{background_cds_panel}}, "length_bounds": {{length_bounds}} } ``` ## Rules - Translate the target and all comparison CDSs under NCBI genetic code 11, stopping each at its first in-frame stop. - Digest every protein completely with trypsin, with zero missed cleavage sites. - A peptide qualifies only if its sequence occurs in the target digest and in none of the comparison digests. Ignore how many copies occur. - The peptide length must be between both inclusive `length_bounds`. ## Selection From eligible target-only peptide sequences, choose the shortest, then the ASCII-lexicographically smallest sequence. Multiple copies of one sequence do not create separate choices. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Trypsin cleavage specificity Trypsin cleaves after K or R, except when the next residue is P. The K or R remains on the upstream peptide. ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. If no target-only peptide meets the length limits, return: ```text NONE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Choose a peptide that distinguishes the target protein from every protein in the comparison panel. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "target_cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "background_cds_panel": ["ATG AAA GCA GTT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA"], "length_bounds": [4, 10] } ``` ## Rules - Translate the target and all comparison CDSs under NCBI genetic code 11, stopping each at its first in-frame stop. - Digest every protein completely with trypsin, with zero missed cleavage sites. - A peptide qualifies only if its sequence occurs in the target digest and in none of the comparison digests. Ignore how many copies occur. - The peptide length must be between both inclusive `length_bounds`. ## Selection From eligible target-only peptide sequences, choose the shortest, then the ASCII-lexicographically smallest sequence. Multiple copies of one sequence do not create separate choices. ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. If no target-only peptide meets the length limits, return: ```text NONE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: AIFVLK` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL is the target. The comparison is a designed I4V variant (codon 4 ATT→GTT), not a second natural accession. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Choose a peptide that distinguishes the target protein from every protein in the comparison panel. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "target_cds": "{{target_cds}}", "background_cds_panel": {{background_cds_panel}}, "length_bounds": {{length_bounds}} } ``` ## Rules - Translate the target and all comparison CDSs under NCBI genetic code 11, stopping each at its first in-frame stop. - Digest every protein completely with trypsin, with zero missed cleavage sites. - A peptide qualifies only if its sequence occurs in the target digest and in none of the comparison digests. Ignore how many copies occur. - The peptide length must be between both inclusive `length_bounds`. ## Selection From eligible target-only peptide sequences, choose the shortest, then the ASCII-lexicographically smallest sequence. Multiple copies of one sequence do not create separate choices. ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. If no target-only peptide meets the length limits, return: ```text NONE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. Selection From eligible target-only peptide sequences, choose the shortest, then the ASCII-lexicographically smallest sequence. Multiple copies of one sequence do not create separate choices. If no target-only peptide meets the length limits, return: NONE Formal fields and scoring Inputs `target_digest` ordered peptide records derived from target CDS `background_digests` peptide records derived from comparison CDSs `length_bounds` inclusive peptide-length range Rules & boundaries Target peptide sequence must be absent from every background digest. Choose shortest eligible peptide, then lexicographically smallest. Sequence uniqueness only; no detectability claim. From eligible target-only peptide sequences, choose the shortest, then the ASCII-lexicographically smallest sequence. Multiple copies of one sequence do not create separate choices. Scoring `exact_match` · complete parsed answer Reference blocks Annotated CDS initiation, NCBI genetic code 11: internal DNA codons, Trypsin cleavage specificity Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "protein.diagnostic_peptide", "output_format": "scalar", "scorer": "exact_match", "family": "protein", "domain": "protein", "objective": "none", "selection_rule": "From eligible target-only peptide sequences, choose the shortest, then the ASCII-lexicographically smallest sequence. Multiple copies of one sequence do not create separate choices." } Used in: From CDS panel to a diagnostic peptide
Calculate model charge at a pH`protein.charge_at_ph`	Translate a CDS and estimate its protein’s net charge at a specified pH using supplied ionization parameters.
Calculate model charge at a pH Some amino-acid side chains and the two ends of a protein gain or lose protons as pH changes. The question supplies a pH and pKa values, which determine the fractions of sites treated as charged in the declared calculation. Count each relevant acidic or basic side chain and both free termini, calculate its contribution with the Henderson–Hasselbalch relation, and add the contributions. The model treats sites independently; it does not account for a folded protein changing a site’s environment. Practical context Estimated protein charge helps choose a pH and resin type for ion-exchange purification, which separates molecules through charge-dependent binding. This calculation supplies a first estimate from the sequence; purification methods then test suitable buffers and binding conditions. Method reference: Cytiva ion-exchanger selection guide. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `ph`: `7.0`; `pka`: supplied table shown in the prompt. Procedure. At pH 7, add the fractional positive charges of the N terminus, two Lys residues, and one Arg, then subtract the fractional negative C-terminal charge. Use the supplied pKa values and round the total once. Final answer: 2.9095 Checking the answer. The net charge must match the supplied-pKa model at the requested pH and precision. Easy example · references supplied Prompt ## Task Calculate the encoded protein’s net charge at `ph`, using the independent-site Henderson–Hasselbalch model and supplied `pka` values. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ph": 7.0, "pka": { "N_term": 8.0, "C_term": 4.0, "K": 10.5, "R": 12.5, "H": 6.0, "D": 3.9, "E": 4.1, "C": 8.3, "Y": 10.1 } } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - Use an unmodified linear protein with both termini free, and the supplied pKa for each site in the independent-site Henderson–Hasselbalch calculation. - Keep the initiating residue. Assume no disulfide bonds or other modifications. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Ionization and net charge A basic site contributes +1/(1+10^(pH−pKa)); an acidic site contributes −1/(1+10^(pKa−pH)). Basic sites: the N terminus and K, R, H side chains. Acidic sites: the C terminus and D, E, C, Y side chains. Count each terminus once and each listed side chain once per residue occurrence; sum their contributions. ## Answer - Return the net charge as a decimal number with 4 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 2.9095` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate the encoded protein’s net charge at `ph`, using the independent-site Henderson–Hasselbalch model and supplied `pka` values. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "ph": {{ph}}, "pka": {{pka}} } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - Use an unmodified linear protein with both termini free, and the supplied pKa for each site in the independent-site Henderson–Hasselbalch calculation. - Keep the initiating residue. Assume no disulfide bonds or other modifications. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Ionization and net charge A basic site contributes +1/(1+10^(pH−pKa)); an acidic site contributes −1/(1+10^(pKa−pH)). Basic sites: the N terminus and K, R, H side chains. Acidic sites: the C terminus and D, E, C, Y side chains. Count each terminus once and each listed side chain once per residue occurrence; sum their contributions. ## Answer - Return the net charge as a decimal number with 4 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Calculate the encoded protein’s net charge at `ph`, using the independent-site Henderson–Hasselbalch model and supplied `pka` values. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ph": 7.0, "pka": { "N_term": 8.0, "C_term": 4.0, "K": 10.5, "R": 12.5, "H": 6.0, "D": 3.9, "E": 4.1, "C": 8.3, "Y": 10.1 } } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - Use an unmodified linear protein with both termini free, and the supplied pKa for each site in the independent-site Henderson–Hasselbalch calculation. - Keep the initiating residue. Assume no disulfide bonds or other modifications. ## Answer - Return the net charge as a decimal number with 4 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 2.9095` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate the encoded protein’s net charge at `ph`, using the independent-site Henderson–Hasselbalch model and supplied `pka` values. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "ph": {{ph}}, "pka": {{pka}} } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - Use an unmodified linear protein with both termini free, and the supplied pKa for each site in the independent-site Henderson–Hasselbalch calculation. - Keep the initiating residue. Assume no disulfide bonds or other modifications. ## Answer - Return the net charge as a decimal number with 4 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return the net charge as a decimal number with 4 digits after the decimal point. Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. Formal fields and scoring Inputs `protein` derived protein string `ph` pH value `pka` values for N_term,C_term,K,R,H,D,E,C,Y Rules & boundaries Independent-site Henderson–Hasselbalch model with free termini and supplied pKa values. Round to 4 decimals. Scoring `exact_match` · complete parsed answer Reference blocks Annotated CDS initiation, NCBI genetic code 11: internal DNA codons, Ionization and net charge Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "protein.charge_at_ph", "output_format": "scalar", "scorer": "exact_match", "family": "protein", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: From CDS to defined protein measurements
Calculate the isoelectric point (pI)`protein.isoelectric_point`	Translate a CDS, then find the pH where the protein’s calculated positive and negative charges cancel.
Calculate the isoelectric point (pI) The isoelectric point (pI) is the pH at which a protein’s positive and negative charges balance, giving a total charge of zero. The inputs are a CDS and pKa values for the groups that can carry charge. A group’s pKa is the pH at which its calculated contribution is half of its fully charged value. The charge calculation is a fixed arithmetic approximation: treat each group separately and add its contribution. Count one N terminus, one C terminus, and each occurrence of the listed amino-acid side chains. Use the supplied pKa for each group in the formulas below; all other side chains contribute zero. Increasing pH decreases this calculated total charge. Each instance is chosen so positive and negative contributions cancel once between pH 0 and pH 14. Those two pH numbers are the initial search bounds. Positive groups: N terminus and K, R, H side chains: `+1 / (1 + 10^(pH − pKa))`. Negative groups: C terminus and D, E, C, Y side chains: `−1 / (1 + 10^(pKa − pH))`. Simple algorithm: repeatedly halve the pH range Translate the CDS and count the charge-carrying groups in that protein, including both free ends. Start with a lower pH of 0 and an upper pH of 14. Calculate total charge at their midpoint, initially pH 7. If the total is positive, the zero-charge pH is higher: replace the lower pH with the midpoint. If negative, it is lower: replace the upper pH. If the total is exactly zero, that pH is the answer. Repeat the midpoint calculation on the remaining pH range until its width is less than 0.001. Report that range’s midpoint rounded to two decimal places. Practical context A protein’s pI helps select conditions for ion-exchange purification and isoelectric focusing. Together with charge at a chosen pH, it indicates which side of the charge transition an operating condition lies on, before testing the behavior experimentally. Method reference: Cytiva ion-exchanger selection guide. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `pka`: supplied table shown in the prompt. Procedure. Evaluate the same charge sum while varying pH from 0 to 14. Bisect that range, keeping the side containing zero charge after each step, and round the pH of the zero crossing to two decimal places. Final answer: 11.67 Checking the answer. Find the unique zero-charge pH under the supplied pKa values and formulas, then round it to two decimal places. That rounded value is the reference answer. Easy example · references supplied Prompt ## Task Calculate the encoded protein’s theoretical pI using the independent-site Henderson–Hasselbalch model and supplied `pka` values. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "pka": { "N_term": 8.0, "C_term": 4.0, "K": 10.5, "R": 12.5, "H": 6.0, "D": 3.9, "E": 4.1, "C": 8.3, "Y": 10.1 } } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - Use the supplied pKa values in the independent-site Henderson–Hasselbalch calculation, including both free termini. - Find the pH at which the total charge is zero, searching from pH 0 to 14. The instance has one such value. - Keep the initiating residue and assume no modifications or disulfide bonds. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Ionization and net charge A basic site contributes +1/(1+10^(pH−pKa)); an acidic site contributes −1/(1+10^(pKa−pH)). Basic sites: the N terminus and K, R, H side chains. Acidic sites: the C terminus and D, E, C, Y side chains. Count each terminus once and each listed side chain once per residue occurrence; sum their contributions. ## Answer - Return the pH as a decimal number with 2 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 11.67` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate the encoded protein’s theoretical pI using the independent-site Henderson–Hasselbalch model and supplied `pka` values. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "pka": {{pka}} } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - Use the supplied pKa values in the independent-site Henderson–Hasselbalch calculation, including both free termini. - Find the pH at which the total charge is zero, searching from pH 0 to 14. The instance has one such value. - Keep the initiating residue and assume no modifications or disulfide bonds. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Ionization and net charge A basic site contributes +1/(1+10^(pH−pKa)); an acidic site contributes −1/(1+10^(pKa−pH)). Basic sites: the N terminus and K, R, H side chains. Acidic sites: the C terminus and D, E, C, Y side chains. Count each terminus once and each listed side chain once per residue occurrence; sum their contributions. ## Answer - Return the pH as a decimal number with 2 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Calculate the encoded protein’s theoretical pI using the independent-site Henderson–Hasselbalch model and supplied `pka` values. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "pka": { "N_term": 8.0, "C_term": 4.0, "K": 10.5, "R": 12.5, "H": 6.0, "D": 3.9, "E": 4.1, "C": 8.3, "Y": 10.1 } } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - Use the supplied pKa values in the independent-site Henderson–Hasselbalch calculation, including both free termini. - Find the pH at which the total charge is zero, searching from pH 0 to 14. The instance has one such value. - Keep the initiating residue and assume no modifications or disulfide bonds. ## Answer - Return the pH as a decimal number with 2 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 11.67` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate the encoded protein’s theoretical pI using the independent-site Henderson–Hasselbalch model and supplied `pka` values. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "pka": {{pka}} } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - Use the supplied pKa values in the independent-site Henderson–Hasselbalch calculation, including both free termini. - Find the pH at which the total charge is zero, searching from pH 0 to 14. The instance has one such value. - Keep the initiating residue and assume no modifications or disulfide bonds. ## Answer - Return the pH as a decimal number with 2 digits after the decimal point. - Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return the pH as a decimal number with 2 digits after the decimal point. Round once at the end using unrounded intermediate values and the declared constants. Round exact halfway cases away from zero; rounded negative zero becomes zero. Formal fields and scoring Inputs `protein` derived protein string `pka` values for N_term,C_term,K,R,H,D,E,C,Y Rules & boundaries Sum the supplied-pKa charge contributions for all listed side chains and both free termini. Find the pH with total charge zero between pH 0 and pH 14; each instance has one such pH. Return the unique root rounded to two decimals. Scoring `exact_match` · complete parsed answer Reference blocks Annotated CDS initiation, NCBI genetic code 11: internal DNA codons, Ionization and net charge Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "protein.isoelectric_point", "output_format": "scalar", "scorer": "exact_match", "family": "protein", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: From CDS to defined protein measurements

05 /

Coding RNA analysis

Editing, pairing, cleavage, and sequence targeting.

9 tasks

rna task definitions
Task	Description	Expand details
Apply specified RNA edits`rna.apply_edits`	Transcribe a CDS, apply explicitly listed changes to its coding RNA, and return the edited RNA sequence.
Apply specified RNA edits RNA editing changes RNA after transcription. One real example is APOBEC1-mediated C-to-U editing of apolipoprotein B mRNA: a CAA codon becomes the UAA stop codon, producing a shorter protein. Here the editing positions and replacement bases are supplied, rather than predicted from an enzyme’s recognition preferences. Apply all changes using positions in that original RNA, keeping RNA’s U alphabet. Replacement text may also insert or delete bases. Later edit coordinates are not shifted by earlier changes. Translation of the edited product is a separate component. Role in related tasks When RNA-editing events have been measured or explicitly proposed, their consequences must be evaluated on the edited sequence. Edit-to-protein analysis uses this step to establish the transcript that is translated; edit-to-fragment analysis instead derives a possible RNA sequence-QC readout. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CAA ACT TCC TGA`; `rna_edits`: `[{"replace": [34, 34, "U"]}]`. Procedure. Transcribe the R12Q variant, then replace original RNA base 34 C with U. Codon 12 changes from CAA to UAA, but this task returns the entire edited RNA rather than translating it. Final answer: AUGAAAGCAAUUUUCGUACUGAAAGGUUGGUGGUAAACUUCCUGA Checking the answer. The complete edited RNA must match the specified changes and preserve every unedited base. Easy example · references supplied Prompt ## Task Apply the supplied edits to the CDS’s coding RNA and return the edited RNA sequence. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CAA ACT TCC TGA", "rna_edits": [ { "replace": [34, 34, "U"] } ] } ``` ## Rules - Transcribe the original CDS and apply the edit objects to that RNA, using original RNA coordinates. - Use the U alphabet in inserted or replacement RNA text. Do not translate the edited RNA. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ## Answer - Return one continuous uppercase 5′→3′ RNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: AUGAAAGCAAUUUUCGUACUGAAAGGUUGGUGGUAAACUUCCUGA` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL with R12Q (CGC→CAA), followed by a supplied C→U editing instruction at RNA base 34. This is a constructed editing exercise, not an observed trpL edit. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Apply the supplied edits to the CDS’s coding RNA and return the edited RNA sequence. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "rna_edits": {{rna_edits}} } ``` ## Rules - Transcribe the original CDS and apply the edit objects to that RNA, using original RNA coordinates. - Use the U alphabet in inserted or replacement RNA text. Do not translate the edited RNA. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ## Answer - Return one continuous uppercase 5′→3′ RNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Apply the supplied edits to the CDS’s coding RNA and return the edited RNA sequence. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CAA ACT TCC TGA", "rna_edits": [ { "replace": [34, 34, "U"] } ] } ``` ## Rules - Transcribe the original CDS and apply the edit objects to that RNA, using original RNA coordinates. - Use the U alphabet in inserted or replacement RNA text. Do not translate the edited RNA. ## Answer - Return one continuous uppercase 5′→3′ RNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: AUGAAAGCAAUUUUCGUACUGAAAGGUUGGUGGUAAACUUCCUGA` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL with R12Q (CGC→CAA), followed by a supplied C→U editing instruction at RNA base 34. This is a constructed editing exercise, not an observed trpL edit. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Apply the supplied edits to the CDS’s coding RNA and return the edited RNA sequence. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "rna_edits": {{rna_edits}} } ``` ## Rules - Transcribe the original CDS and apply the edit objects to that RNA, using original RNA coordinates. - Use the U alphabet in inserted or replacement RNA text. Do not translate the edited RNA. ## Answer - Return one continuous uppercase 5′→3′ RNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase 5′→3′ RNA sequence. An empty sequence is the token EMPTY_SEQUENCE. Formal fields and scoring Inputs `rna` coding RNA derived from CDS `rna_edits` nonoverlapping original-coordinate RNA replacements/insertions Rules & boundaries Explicit changes only; no inference of editing propensity. Replacement alphabet is A,C,G,U. Scoring `exact_match` · complete parsed answer Reference blocks Coding DNA to RNA Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "rna.apply_edits", "output_format": "scalar", "scorer": "exact_match", "family": "rna", "domain": "rna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: From specified RNA edits to protein From specified RNA edits to RNA fragments Design and verify an RNA-editing-dependent product
Translate coding RNA`rna.translate`	Translate coding RNA derived from a CDS, including RNA that has undergone supplied edits.
Translate coding RNA This component accepts RNA rather than DNA so an edited transcript can be passed directly into translation. In an isolated prompt, the RNA is transcribed from the supplied CDS; in a composition, it can be the output of the RNA-editing task. Read RNA codons from the established first base using the named genetic code, with U in place of DNA’s T. Stop at the first in-frame stop and omit that stop from the answer. Do not silently switch back to translating the original DNA. Role in related tasks The useful distinction from direct DNA translation appears in the RNA-editing composition: the protein must be derived from the edited coding RNA, which can differ from the product predicted from the original DNA. This task supplies that downstream translation step. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`. Procedure. Replace coding-strand T with U and read codons from base 1 under code 11. The coding RNA gives the same MKAIFVLKGWWRTS protein as the DNA translation task. Final answer: MKAIFVLKGWWRTS Checking the answer. The returned protein must come from the actual RNA input, including its edits and stopping point. Easy example · references supplied Prompt ## Task Translate the CDS’s coding RNA using NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Start translation at the RNA’s first base, using NCBI genetic code 11. Do not search for a different reading origin. - End at the first in-frame stop and omit the stop symbol. Ignore an incomplete trailing codon. - Apply the initiator rule when the first codon is an allowed initiator. Otherwise translate that first codon with its ordinary internal meaning. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Initiation of coding RNA At the first RNA codon, UUG, CUG, AUU, AUC, AUA, AUG, and GUG encode initiating M under NCBI genetic code 11. If that first codon is not an allowed initiator, use its ordinary internal assignment. ### NCBI genetic code 11: internal RNA codons Each internal RNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text UUU=F UUC=F UUA=L UUG=L UCU=S UCC=S UCA=S UCG=S UAU=Y UAC=Y UAA=STOP UAG=STOP UGU=C UGC=C UGA=STOP UGG=W CUU=L CUC=L CUA=L CUG=L CCU=P CCC=P CCA=P CCG=P CAU=H CAC=H CAA=Q CAG=Q CGU=R CGC=R CGA=R CGG=R AUU=I AUC=I AUA=I AUG=M ACU=T ACC=T ACA=T ACG=T AAU=N AAC=N AAA=K AAG=K AGU=S AGC=S AGA=R AGG=R GUU=V GUC=V GUA=V GUG=V GCU=A GCC=A GCA=A GCG=A GAU=D GAC=D GAA=E GAG=E GGU=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: MKAIFVLKGWWRTS` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Translate the CDS’s coding RNA using NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Start translation at the RNA’s first base, using NCBI genetic code 11. Do not search for a different reading origin. - End at the first in-frame stop and omit the stop symbol. Ignore an incomplete trailing codon. - Apply the initiator rule when the first codon is an allowed initiator. Otherwise translate that first codon with its ordinary internal meaning. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Initiation of coding RNA At the first RNA codon, UUG, CUG, AUU, AUC, AUA, AUG, and GUG encode initiating M under NCBI genetic code 11. If that first codon is not an allowed initiator, use its ordinary internal assignment. ### NCBI genetic code 11: internal RNA codons Each internal RNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text UUU=F UUC=F UUA=L UUG=L UCU=S UCC=S UCA=S UCG=S UAU=Y UAC=Y UAA=STOP UAG=STOP UGU=C UGC=C UGA=STOP UGG=W CUU=L CUC=L CUA=L CUG=L CCU=P CCC=P CCA=P CCG=P CAU=H CAC=H CAA=Q CAG=Q CGU=R CGC=R CGA=R CGG=R AUU=I AUC=I AUA=I AUG=M ACU=T ACC=T ACA=T ACG=T AAU=N AAC=N AAA=K AAG=K AGU=S AGC=S AGA=R AGG=R GUU=V GUC=V GUA=V GUG=V GCU=A GCC=A GCA=A GCG=A GAU=D GAC=D GAA=E GAG=E GGU=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Translate the CDS’s coding RNA using NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Start translation at the RNA’s first base, using NCBI genetic code 11. Do not search for a different reading origin. - End at the first in-frame stop and omit the stop symbol. Ignore an incomplete trailing codon. - Apply the initiator rule when the first codon is an allowed initiator. Otherwise translate that first codon with its ordinary internal meaning. ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: MKAIFVLKGWWRTS` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Translate the CDS’s coding RNA using NCBI genetic code 11. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Start translation at the RNA’s first base, using NCBI genetic code 11. Do not search for a different reading origin. - End at the first in-frame stop and omit the stop symbol. Ignore an incomplete trailing codon. - Apply the initiator rule when the first codon is an allowed initiator. Otherwise translate that first codon with its ordinary internal meaning. ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. An empty sequence is the token EMPTY_SEQUENCE. Formal fields and scoring Inputs `rna` coding RNA derived from CDS, possibly edited Rules & boundaries NCBI 11 with U replacing T in codons; first-stop termination; initiator rule applies at original first codon if valid. Scoring `exact_match` · complete parsed answer Reference blocks Coding DNA to RNA, Initiation of coding RNA, NCBI genetic code 11: internal RNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "rna.translate", "output_format": "scalar", "scorer": "exact_match", "family": "rna", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: From specified RNA edits to protein Design and verify an RNA-editing-dependent product
Check specified RNA pairs`rna.pair_compatibility`	Given a CDS-derived RNA and specified paired positions, identify pairs whose bases cannot form an allowed pair.
Check specified RNA pairs Complementary stretches of one RNA can bind each other, forming paired stems and unpaired loops. A–U and C–G are the canonical base-pair types. The supplied position pairs state which contacts to check; the task does not discover the fold. Read the two bases at each listed pair and test their compatibility. Real RNA can also form G–U wobble pairs, but this benchmark deliberately permits only A–U and C–G in either direction. Return the incompatible position pairs only. Role in related tasks This is a local consistency check after changing bases in a proposed RNA stem. Pair repair with protein verification uses it to test whether a synonymous redesign restores the requested contacts. The structure is supplied, so the check does not establish that those contacts form experimentally. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `pairs`: `[[4, 13], [5, 12], [7, 15], [25, 33]]`. Procedure. Look up both RNA bases for each supplied pair. Pairs (4,13) and (5,12) are A–U; (7,15) is G–C; (25,33) is G–G and is the sole incompatible pair. Final answer: [[25, 33]] Checking the answer. The answer must identify every incompatible supplied pair. It does not assert that the RNA actually adopts the supplied structure. Easy example · references supplied Prompt ## Task Identify the listed RNA position pairs that cannot form canonical Watson–Crick pairs. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "pairs": [ [4, 13], [5, 12], [7, 15], [25, 33] ] } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Check each pair of positions in the supplied map against canonical Watson–Crick pairing. Do not allow wobble pairs. - Report only incompatible listed pairs. Do not infer additional pairs or predict a new structure. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ## Answer - Return a list of [first_position, second_position] integer pairs in the coding RNA, with the smaller 1-based position first. - Include each qualifying pair once, sorted by its first and then second position. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: [[25, 33]]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Pairs are supplied test annotations, not an experimentally determined RNA structure. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Identify the listed RNA position pairs that cannot form canonical Watson–Crick pairs. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pairs": {{pairs}} } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Check each pair of positions in the supplied map against canonical Watson–Crick pairing. Do not allow wobble pairs. - Report only incompatible listed pairs. Do not infer additional pairs or predict a new structure. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ## Answer - Return a list of [first_position, second_position] integer pairs in the coding RNA, with the smaller 1-based position first. - Include each qualifying pair once, sorted by its first and then second position. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Identify the listed RNA position pairs that cannot form canonical Watson–Crick pairs. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "pairs": [ [4, 13], [5, 12], [7, 15], [25, 33] ] } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Check each pair of positions in the supplied map against canonical Watson–Crick pairing. Do not allow wobble pairs. - Report only incompatible listed pairs. Do not infer additional pairs or predict a new structure. ## Answer - Return a list of [first_position, second_position] integer pairs in the coding RNA, with the smaller 1-based position first. - Include each qualifying pair once, sorted by its first and then second position. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: [[25, 33]]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Pairs are supplied test annotations, not an experimentally determined RNA structure. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Identify the listed RNA position pairs that cannot form canonical Watson–Crick pairs. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pairs": {{pairs}} } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Check each pair of positions in the supplied map against canonical Watson–Crick pairing. Do not allow wobble pairs. - Report only incompatible listed pairs. Do not infer additional pairs or predict a new structure. ## Answer - Return a list of [first_position, second_position] integer pairs in the coding RNA, with the smaller 1-based position first. - Include each qualifying pair once, sorted by its first and then second position. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return a list of [first_position, second_position] integer pairs in the coding RNA, with the smaller 1-based position first. Include each qualifying pair once, sorted by its first and then second position. An empty result is []. Formal fields and scoring Inputs `rna` coding RNA `pairs` unique position pairs i<j; no position used twice Rules & boundaries Canonical Watson–Crick pairs only; map is input, not a folding prediction. Scoring `exact_match` · complete parsed answer Reference blocks Coding DNA to RNA, Canonical RNA base pairs Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "rna.pair_compatibility", "output_format": "scalar", "scorer": "exact_match", "family": "rna", "domain": "rna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: Repair RNA pairs and verify the protein
Decode a pairing map`rna.dotbracket_to_pairs`	Convert a supplied dot-bracket annotation for coding RNA into explicit pairs of nucleotide positions.
Decode a pairing map Dot-bracket notation describes an RNA pairing map. A dot marks an unpaired base; matching opening and closing parentheses identify a pair. The annotation has the same length as the RNA. Match the parentheses and report their 1-based positions. Nested pairs are allowed, but this task excludes crossing pairs called pseudoknots. This is notation interpretation rather than prediction of a structure from the sequence. Role in related tasks Dot-bracket notation is a common output of RNA structure software. Converting it to coordinate pairs makes the annotation usable in explicit sequence checks, such as pair compatibility and the simplified structure-dependent cleavage model. Method reference: ViennaRNA. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `dot_bracket`: `...(((....)))...........((.......).).........`. Procedure. Read the 45-character annotation left to right. Push each opening-parenthesis position onto a stack; pair each closing parenthesis with the most recent unmatched opening position. Sort the five resulting pairs. Final answer: [[4, 13], [5, 12], [6, 11], [25, 36], [26, 34]] Checking the answer. Every parenthesis pair must become exactly one correct coordinate pair. Easy example Prompt ## Task Convert the supplied RNA dot-bracket annotation into a list of paired nucleotide positions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "dot_bracket": "...(((....)))...........((.......).)........." } ``` ## Rules - The supplied annotation is balanced and contains only `(`, `)`, and `.`. It has one character per RNA base. - Each closing parenthesis pairs with the most recent unmatched opening parenthesis. Dots represent unpaired positions. ## Answer - Return a list of [first_position, second_position] integer pairs in the coding RNA, with the smaller 1-based position first. - Include every pair once, sorted by its first and then second position. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: [[4, 13], [5, 12], [6, 11], [25, 36], [26, 34]]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Dot-bracket is a supplied illustrative noncrossing annotation, not a measured structure. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Convert the supplied RNA dot-bracket annotation into a list of paired nucleotide positions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "dot_bracket": "{{dot_bracket}}" } ``` ## Rules - The supplied annotation is balanced and contains only `(`, `)`, and `.`. It has one character per RNA base. - Each closing parenthesis pairs with the most recent unmatched opening parenthesis. Dots represent unpaired positions. ## Answer - Return a list of [first_position, second_position] integer pairs in the coding RNA, with the smaller 1-based position first. - Include every pair once, sorted by its first and then second position. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example Prompt ## Task Convert the supplied RNA dot-bracket annotation into a list of paired nucleotide positions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "dot_bracket": "...(((....)))...........((.......).)........." } ``` ## Rules - The supplied annotation is balanced and contains only `(`, `)`, and `.`. It has one character per RNA base. - Each closing parenthesis pairs with the most recent unmatched opening parenthesis. Dots represent unpaired positions. ## Answer - Return a list of [first_position, second_position] integer pairs in the coding RNA, with the smaller 1-based position first. - Include every pair once, sorted by its first and then second position. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: [[4, 13], [5, 12], [6, 11], [25, 36], [26, 34]]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Dot-bracket is a supplied illustrative noncrossing annotation, not a measured structure. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Convert the supplied RNA dot-bracket annotation into a list of paired nucleotide positions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "dot_bracket": "{{dot_bracket}}" } ``` ## Rules - The supplied annotation is balanced and contains only `(`, `)`, and `.`. It has one character per RNA base. - Each closing parenthesis pairs with the most recent unmatched opening parenthesis. Dots represent unpaired positions. ## Answer - Return a list of [first_position, second_position] integer pairs in the coding RNA, with the smaller 1-based position first. - Include every pair once, sorted by its first and then second position. An empty result is []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return a list of [first_position, second_position] integer pairs in the coding RNA, with the smaller 1-based position first. Include every pair once, sorted by its first and then second position. An empty result is []. Formal fields and scoring Inputs `rna` coding RNA `dot_bracket` balanced .() string of same length; no pseudoknots Rules & boundaries Match each closing parenthesis to most recent unmatched opening parenthesis; ignore dots. Scoring `exact_match` · complete parsed answer Reference blocks None required. Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "rna.dotbracket_to_pairs", "output_format": "scalar", "scorer": "exact_match", "family": "rna", "domain": "rna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: From a pairing annotation to eligible cuts
Digest coding RNA with RNase T1 rules`rna.t1_digest`	Transcribe a CDS and split its RNA into fragments using an RNase T1 rule, optionally restricted by a supplied pairing map.
Digest coding RNA with RNase T1 rules RNase T1 is a real guanosine-specific RNA-cutting enzyme originally obtained from the fungus Aspergillus oryzae. It cleaves the RNA backbone immediately after G, on the G’s 3′ side. The G therefore remains at the end of the upstream fragment. Choosing G reflects this enzyme’s biochemical specificity. RNase T1 acts preferentially on single-stranded RNA. The question either treats every G as eligible or uses a supplied pairing map to allow cuts only at unpaired G bases. Complete cleavage and that fixed accessibility mask are benchmark simplifications. Scan for eligible cuts, split the sequence, and return the fragments in sequence order with their original RNA coordinates. When a pairing map is supplied, keep it fixed rather than refolding after each cut. Practical context RNase fragment mapping is used in RNA sequence characterization, including mRNA quality control. This task supplies the simple sequence-to-fragments step. Practical mapping combines digestion with mass spectrometry and may use complementary or partial digests to resolve fragments that a single complete digest cannot distinguish. Method reference: RNA and mRNA sequence mapping by mass spectrometry. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `accessibility`: `all`; `pairs`: `[]`. Procedure. Transcribe all 45 bases and cut immediately after each G, the specificity of RNase T1 in this benchmark. Every position is accessible; retain every nonempty fragment and its original RNA coordinates. Final answer: [ { "start": 1, "end": 3, "fragment": "AUG" }, { "start": 4, "end": 7, "fragment": "AAAG" }, { "start": 8, "end": 16, "fragment": "CAAUUUUCG" }, { "start": 17, "end": 21, "fragment": "UACUG" }, { "start": 22, "end": 25, "fragment": "AAAG" }, { "start": 26, "end": 26, "fragment": "G" }, { "start": 27, "end": 29, "fragment": "UUG" }, { "start": 30, "end": 30, "fragment": "G" }, { "start": 31, "end": 32, "fragment": "UG" }, { "start": 33, "end": 33, "fragment": "G" }, { "start": 34, "end": 35, "fragment": "CG" }, { "start": 36, "end": 44, "fragment": "CACUUCCUG" }, { "start": 45, "end": 45, "fragment": "A" } ] Checking the answer. The fragments must reflect both the cleavage rule and the requested accessibility setting. Experimental digestion efficiency is not part of the score. Easy example · references supplied Prompt ## Task Digest the coding RNA with RNase T1, allowing cuts according to `accessibility` and the supplied pairing map. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "accessibility": "all", "pairs": [] } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Apply the RNase T1 specificity rule with zero missed eligible cuts. - With `accessibility=all`, ignore the pairing map when deciding where the enzyme can cut. - With `accessibility=unpaired_only`, allow a cut only if the nucleotide recognized by the enzyme appears in no pair in the supplied map. Keep that map fixed after every cut. - Return every nonempty fragment, including any uncut fragment at the end. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNase T1 cleavage specificity RNase T1 cleaves RNA after G. That G remains on the upstream fragment. ## Answer - `start` and `end` are 1-based inclusive positions in the supplied CDS-derived RNA; `fragment` is the continuous uppercase RNA sequence. - Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer Final answer: [ { "start": 1, "end": 3, "fragment": "AUG" }, { "start": 4, "end": 7, "fragment": "AAAG" }, { "start": 8, "end": 16, "fragment": "CAAUUUUCG" }, { "start": 17, "end": 21, "fragment": "UACUG" }, { "start": 22, "end": 25, "fragment": "AAAG" }, { "start": 26, "end": 26, "fragment": "G" }, { "start": 27, "end": 29, "fragment": "UUG" }, { "start": 30, "end": 30, "fragment": "G" }, { "start": 31, "end": 32, "fragment": "UG" }, { "start": 33, "end": 33, "fragment": "G" }, { "start": 34, "end": 35, "fragment": "CG" }, { "start": 36, "end": 44, "fragment": "CACUUCCUG" }, { "start": 45, "end": 45, "fragment": "A" } ] Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. All positions are accessible in this example; the separate structure-gated digest uses a supplied pairing map. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Digest the coding RNA with RNase T1, allowing cuts according to `accessibility` and the supplied pairing map. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "accessibility": "{{accessibility}}", "pairs": {{pairs}} } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Apply the RNase T1 specificity rule with zero missed eligible cuts. - With `accessibility=all`, ignore the pairing map when deciding where the enzyme can cut. - With `accessibility=unpaired_only`, allow a cut only if the nucleotide recognized by the enzyme appears in no pair in the supplied map. Keep that map fixed after every cut. - Return every nonempty fragment, including any uncut fragment at the end. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNase T1 cleavage specificity RNase T1 cleaves RNA after G. That G remains on the upstream fragment. ## Answer - `start` and `end` are 1-based inclusive positions in the supplied CDS-derived RNA; `fragment` is the continuous uppercase RNA sequence. - Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Digest the coding RNA with RNase T1, allowing cuts according to `accessibility` and the supplied pairing map. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "accessibility": "all", "pairs": [] } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Apply the RNase T1 specificity rule with zero missed eligible cuts. - With `accessibility=all`, ignore the pairing map when deciding where the enzyme can cut. - With `accessibility=unpaired_only`, allow a cut only if the nucleotide recognized by the enzyme appears in no pair in the supplied map. Keep that map fixed after every cut. - Return every nonempty fragment, including any uncut fragment at the end. ## Answer - `start` and `end` are 1-based inclusive positions in the supplied CDS-derived RNA; `fragment` is the continuous uppercase RNA sequence. - Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer Final answer: [ { "start": 1, "end": 3, "fragment": "AUG" }, { "start": 4, "end": 7, "fragment": "AAAG" }, { "start": 8, "end": 16, "fragment": "CAAUUUUCG" }, { "start": 17, "end": 21, "fragment": "UACUG" }, { "start": 22, "end": 25, "fragment": "AAAG" }, { "start": 26, "end": 26, "fragment": "G" }, { "start": 27, "end": 29, "fragment": "UUG" }, { "start": 30, "end": 30, "fragment": "G" }, { "start": 31, "end": 32, "fragment": "UG" }, { "start": 33, "end": 33, "fragment": "G" }, { "start": 34, "end": 35, "fragment": "CG" }, { "start": 36, "end": 44, "fragment": "CACUUCCUG" }, { "start": 45, "end": 45, "fragment": "A" } ] Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. All positions are accessible in this example; the separate structure-gated digest uses a supplied pairing map. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Digest the coding RNA with RNase T1, allowing cuts according to `accessibility` and the supplied pairing map. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "accessibility": "{{accessibility}}", "pairs": {{pairs}} } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Apply the RNase T1 specificity rule with zero missed eligible cuts. - With `accessibility=all`, ignore the pairing map when deciding where the enzyme can cut. - With `accessibility=unpaired_only`, allow a cut only if the nucleotide recognized by the enzyme appears in no pair in the supplied map. Keep that map fixed after every cut. - Return every nonempty fragment, including any uncut fragment at the end. ## Answer - `start` and `end` are 1-based inclusive positions in the supplied CDS-derived RNA; `fragment` is the continuous uppercase RNA sequence. - Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `start` and `end` are 1-based inclusive positions in the supplied CDS-derived RNA; `fragment` is the continuous uppercase RNA sequence. Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] Formal fields and scoring Inputs `rna` coding RNA `accessibility` all \| unpaired_only `pairs` fixed pair map; [] if all Rules & boundaries Rule t1_after_G_v1. In unpaired_only mode, G must be absent from the fixed map. Complete digestion; no cyclic fragments. Scoring `exact_match` · complete parsed answer Reference blocks Coding DNA to RNA, RNase T1 cleavage specificity Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "rna.t1_digest", "output_format": "json", "scorer": "exact_match", "family": "rna", "domain": "rna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: From specified RNA edits to RNA fragments From a pairing annotation to eligible cuts
Match an RNA guide`rna.guide_hits`	Given an RNA guide and a CDS-derived RNA, find target intervals complementary to the guide within an allowed mismatch count.
Match an RNA guide Guide-directed recognition is real biology: Argonaute proteins use guide RNAs to recognize complementary mRNA in RNA silencing. The two RNAs pair in opposite directions, so compare target windows with the reverse complement of the supplied 5′→3′ guide. Examine every target window as long as the guide. Count unequal positions and retain windows at or below the given maximum. This is a simplified sequence screen: real guide systems distinguish mismatch positions and structural contexts, while this task uses a uniform mismatch count with no gaps or wobble. Role in related tasks Complementary-hit screening helps identify candidate RNA probes or target-site changes. Probe selection uses the hits for sequence discrimination, and guide-escape design uses them to disrupt a specified interaction. Broader oligonucleotide design methods also model binding thermodynamics and competing structures. Method reference: NUPACK documentation. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `guide`: `ACCUUUCAGUACGAAAAUUGC`; `max_mismatches`: `1`. Procedure. Reverse-complement the 21-base guide, compare it with each of the 25 equal-length RNA windows, and count mismatches. The designed matching window spans bases 7–27. Final answer: [ { "target_start": 7, "target_end": 27, "mismatches": 0 } ] Checking the answer. Return all qualifying intervals and their mismatch counts. Sequence compatibility does not establish cellular guide activity. Easy example · references supplied Prompt ## Task Find every coding-RNA window that matches the RNA guide within `max_mismatches`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "guide": "ACCUUUCAGUACGAAAAUUGC", "max_mismatches": 1 } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Compare the reverse complement of the guide with every equal-length window of the coding RNA. The guide is supplied 5′→3′. - For each window, count positions with different bases. Keep the window if that count is at most `max_mismatches`, including equality. - Use ungapped comparisons and exclude wobble pairing. Include overlapping windows. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNA reverse complement RNA complementary bases are A↔U and C↔G. Complement each base, then reverse the sequence to write its complementary partner 5′→3′. ## Answer - `target_start` and `target_end` are 1-based inclusive coding-RNA coordinates. `mismatches` is the integer Hamming distance between that window and the guide’s reverse complement. - List every qualifying window in increasing target_start order, including overlaps. An empty result is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "target_start": <positive integer>, "target_end": <positive integer>, "mismatches": <nonnegative integer> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: [ { "target_start": 7, "target_end": 27, "mismatches": 0 } ]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The 21-base guide is designed as the reverse complement of coding-RNA bases 7–27; it is not an experimentally validated reagent. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find every coding-RNA window that matches the RNA guide within `max_mismatches`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "guide": "{{guide}}", "max_mismatches": {{max_mismatches}} } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Compare the reverse complement of the guide with every equal-length window of the coding RNA. The guide is supplied 5′→3′. - For each window, count positions with different bases. Keep the window if that count is at most `max_mismatches`, including equality. - Use ungapped comparisons and exclude wobble pairing. Include overlapping windows. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNA reverse complement RNA complementary bases are A↔U and C↔G. Complement each base, then reverse the sequence to write its complementary partner 5′→3′. ## Answer - `target_start` and `target_end` are 1-based inclusive coding-RNA coordinates. `mismatches` is the integer Hamming distance between that window and the guide’s reverse complement. - List every qualifying window in increasing target_start order, including overlaps. An empty result is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "target_start": <positive integer>, "target_end": <positive integer>, "mismatches": <nonnegative integer> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Find every coding-RNA window that matches the RNA guide within `max_mismatches`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "guide": "ACCUUUCAGUACGAAAAUUGC", "max_mismatches": 1 } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Compare the reverse complement of the guide with every equal-length window of the coding RNA. The guide is supplied 5′→3′. - For each window, count positions with different bases. Keep the window if that count is at most `max_mismatches`, including equality. - Use ungapped comparisons and exclude wobble pairing. Include overlapping windows. ## Answer - `target_start` and `target_end` are 1-based inclusive coding-RNA coordinates. `mismatches` is the integer Hamming distance between that window and the guide’s reverse complement. - List every qualifying window in increasing target_start order, including overlaps. An empty result is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "target_start": <positive integer>, "target_end": <positive integer>, "mismatches": <nonnegative integer> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: [ { "target_start": 7, "target_end": 27, "mismatches": 0 } ]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The 21-base guide is designed as the reverse complement of coding-RNA bases 7–27; it is not an experimentally validated reagent. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find every coding-RNA window that matches the RNA guide within `max_mismatches`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "guide": "{{guide}}", "max_mismatches": {{max_mismatches}} } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Compare the reverse complement of the guide with every equal-length window of the coding RNA. The guide is supplied 5′→3′. - For each window, count positions with different bases. Keep the window if that count is at most `max_mismatches`, including equality. - Use ungapped comparisons and exclude wobble pairing. Include overlapping windows. ## Answer - `target_start` and `target_end` are 1-based inclusive coding-RNA coordinates. `mismatches` is the integer Hamming distance between that window and the guide’s reverse complement. - List every qualifying window in increasing target_start order, including overlaps. An empty result is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "target_start": <positive integer>, "target_end": <positive integer>, "mismatches": <nonnegative integer> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `target_start` and `target_end` are 1-based inclusive coding-RNA coordinates. `mismatches` is the integer Hamming distance between that window and the guide’s reverse complement. List every qualifying window in increasing target_start order, including overlaps. An empty result is []. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. [ { "target_start": <positive integer>, "target_end": <positive integer>, "mismatches": <nonnegative integer> } ] Formal fields and scoring Inputs `rna` coding RNA `guide` RNA guide written 5′→3′ `max_mismatches` nonnegative integer Rules & boundaries Compare every same-length RNA window with reverse complement of guide; Hamming mismatches, no gaps or wobble. No PAM (RNA target); specificity only within supplied sequences. Scoring `exact_match` · complete parsed answer Reference blocks Coding DNA to RNA, RNA reverse complement Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "rna.guide_hits", "output_format": "json", "scorer": "exact_match", "family": "rna", "domain": "rna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: Choose an RNA probe and a matched escape control From CDS collection to a minimal measurement panel
Count coding-RNA patterns`rna.motif_counts`	Transcribe the CDS and count occurrences of specified short RNA sequences, including overlapping occurrences.
Count coding-RNA patterns The question supplies literal RNA patterns to measure, such as a two-base sequence or a longer recognition pattern. No recognition motif needs to be inferred from biological context. Check every possible RNA starting position for each pattern. Include matches crossing codon boundaries, and report a count of zero when a supplied pattern is absent. This gives a simple measurement that synonymous-design tasks can control. Role in related tasks A pattern can be a probe’s exact complementary target sequence; counting it measures how many such candidate sites are present. Probe selection adds comparison RNAs, while motif control changes the count through synonymous recoding. Arbitrary patterns have no biological interpretation until a recognition process or experiment is named. Method reference: NUPACK documentation. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `patterns`: `["UGG", "UGGUGG", "AAA", "CGCG"]`. Procedure. Search separately for each exact RNA pattern, moving one base at a time. Count both adjacent UGG codons and the spanning UGGUGG occurrence; include zero for an absent requested pattern. Final answer: { "CGCG": 0, "AAA": 2, "UGG": 2, "UGGUGG": 1 } Checking the answer. Each supplied pattern must have its exact occurrence count, including overlaps. Easy example · references supplied Prompt ## Task Count overlapping occurrences of these RNA patterns in coding RNA. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "patterns": ["UGG", "UGGUGG", "AAA", "CGCG"] } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Count exact matches on the forward coding RNA, including overlapping occurrences and occurrences spanning codon boundaries. - Include a count of zero for each supplied pattern with no matches. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ## Answer - Map every distinct supplied uppercase RNA pattern to its nonnegative integer count of overlapping forward occurrences in the coding RNA. - Include requested patterns with zero occurrences. Repeated identical input patterns share one key. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<supplied RNA pattern>": <nonnegative integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "CGCG": 0, "AAA": 2, "UGG": 2, "UGGUGG": 1 }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Count overlapping occurrences of these RNA patterns in coding RNA. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "patterns": {{patterns}} } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Count exact matches on the forward coding RNA, including overlapping occurrences and occurrences spanning codon boundaries. - Include a count of zero for each supplied pattern with no matches. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ## Answer - Map every distinct supplied uppercase RNA pattern to its nonnegative integer count of overlapping forward occurrences in the coding RNA. - Include requested patterns with zero occurrences. Repeated identical input patterns share one key. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<supplied RNA pattern>": <nonnegative integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Count overlapping occurrences of these RNA patterns in coding RNA. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "patterns": ["UGG", "UGGUGG", "AAA", "CGCG"] } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Count exact matches on the forward coding RNA, including overlapping occurrences and occurrences spanning codon boundaries. - Include a count of zero for each supplied pattern with no matches. ## Answer - Map every distinct supplied uppercase RNA pattern to its nonnegative integer count of overlapping forward occurrences in the coding RNA. - Include requested patterns with zero occurrences. Repeated identical input patterns share one key. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<supplied RNA pattern>": <nonnegative integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "CGCG": 0, "AAA": 2, "UGG": 2, "UGGUGG": 1 }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Count overlapping occurrences of these RNA patterns in coding RNA. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "patterns": {{patterns}} } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Count exact matches on the forward coding RNA, including overlapping occurrences and occurrences spanning codon boundaries. - Include a count of zero for each supplied pattern with no matches. ## Answer - Map every distinct supplied uppercase RNA pattern to its nonnegative integer count of overlapping forward occurrences in the coding RNA. - Include requested patterns with zero occurrences. Repeated identical input patterns share one key. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<supplied RNA pattern>": <nonnegative integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` Map every distinct supplied uppercase RNA pattern to its nonnegative integer count of overlapping forward occurrences in the coding RNA. Include requested patterns with zero occurrences. Repeated identical input patterns share one key. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "<supplied RNA pattern>": <nonnegative integer count> } Formal fields and scoring Inputs `rna` coding RNA `patterns` nonempty literal RNA strings Rules & boundaries Exact forward matches; count overlaps; include zero counts. No inference of biological binding. Scoring `exact_match` · complete parsed answer Reference blocks Coding DNA to RNA Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "rna.motif_counts", "output_format": "json", "scorer": "exact_match", "family": "rna", "domain": "rna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: Build and audit a controlled synonymous panel
Score a specified RNA structure`rna.structure_score`	Given a coding RNA, a pairing map, and numerical pair scores, calculate the score of that particular structure.
Score a specified RNA structure Both the structure and its scoring table are supplied. The task therefore does not require finding a fold or knowing which structure is most stable in a cell. First check that every listed pair is allowed and that no base is assigned two partners. If valid, add the supplied score for each pair. Return INVALID if the map violates those conditions. The additive score is a declared toy objective, not thermodynamic free energy. Exercise scope RNA base pairing is physical, but the numerical bonuses in this task are supplied exercise scores. Joint RNA-and-codon optimization uses them as an exactly checkable constraint. Practical methods such as LinearDesign use thermodynamic energy models that include stacking and loops, rather than assigning an arbitrary bonus to each isolated pair. Method reference: LinearDesign. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `pairs`: `[[4, 13], [5, 12], [25, 36], [26, 34]]`; `pair_scores`: `{"AU": 2, "UA": 2, "CG": 3, "GC": 3}`. Procedure. Read bases at the four listed contacts. Two A–U contacts score 2 each and two G–C contacts score 3 each. No position is reused; add 2+2+3+3. Final answer: 10 Checking the answer. The validity judgement and, when valid, the total must be exact. Easy example · references supplied Prompt ## Task Calculate the score of the supplied RNA pairing map using `pair_scores`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "pairs": [ [4, 13], [5, 12], [25, 36], [26, 34] ], "pair_scores": { "AU": 2, "UA": 2, "CG": 3, "GC": 3 } } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Use the supplied position pairs as a fixed map. For each pair, look up the score of its two RNA bases in `pair_scores`, then add the scores. - Return `INVALID` if any listed pair is not canonical or if any nucleotide position appears in more than one pair. - Do not optimize the map or predict a different structure. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ## Answer - Return the integer sum of the supplied scores for the listed pairs. If a listed pair is noncanonical or a position is reused, return: ```text INVALID ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 10` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Pair map and integer scores are supplied benchmark inputs, not a folding-energy model or measured structure. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate the score of the supplied RNA pairing map using `pair_scores`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pairs": {{pairs}}, "pair_scores": {{pair_scores}} } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Use the supplied position pairs as a fixed map. For each pair, look up the score of its two RNA bases in `pair_scores`, then add the scores. - Return `INVALID` if any listed pair is not canonical or if any nucleotide position appears in more than one pair. - Do not optimize the map or predict a different structure. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ## Answer - Return the integer sum of the supplied scores for the listed pairs. If a listed pair is noncanonical or a position is reused, return: ```text INVALID ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Calculate the score of the supplied RNA pairing map using `pair_scores`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "pairs": [ [4, 13], [5, 12], [25, 36], [26, 34] ], "pair_scores": { "AU": 2, "UA": 2, "CG": 3, "GC": 3 } } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Use the supplied position pairs as a fixed map. For each pair, look up the score of its two RNA bases in `pair_scores`, then add the scores. - Return `INVALID` if any listed pair is not canonical or if any nucleotide position appears in more than one pair. - Do not optimize the map or predict a different structure. ## Answer - Return the integer sum of the supplied scores for the listed pairs. If a listed pair is noncanonical or a position is reused, return: ```text INVALID ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: 10` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Pair map and integer scores are supplied benchmark inputs, not a folding-energy model or measured structure. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Calculate the score of the supplied RNA pairing map using `pair_scores`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pairs": {{pairs}}, "pair_scores": {{pair_scores}} } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Use the supplied position pairs as a fixed map. For each pair, look up the score of its two RNA bases in `pair_scores`, then add the scores. - Return `INVALID` if any listed pair is not canonical or if any nucleotide position appears in more than one pair. - Do not optimize the map or predict a different structure. ## Answer - Return the integer sum of the supplied scores for the listed pairs. If a listed pair is noncanonical or a position is reused, return: ```text INVALID ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return the integer sum of the supplied scores for the listed pairs. If a listed pair is noncanonical or a position is reused, return: INVALID Formal fields and scoring Inputs `rna` coding RNA `pairs` noncrossing pair map `pair_scores` explicit scores for AU,UA,CG,GC Rules & boundaries wc_additive_v1: sum supplied pair scores; invalid pair or repeated position→INVALID. Not a thermodynamic free energy. Scoring `exact_match` · complete parsed answer Reference blocks Coding DNA to RNA, Canonical RNA base pairs Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "rna.structure_score", "output_format": "scalar", "scorer": "exact_match", "family": "rna", "domain": "rna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: Balance an RNA structural constraint and codon preference
Find a maximum-pair RNA fold`rna.max_pair_fold`	Find a noncrossing arrangement of complementary base pairs in a CDS-derived RNA, with as many pairs as possible.
Find a maximum-pair RNA fold Choose which positions in the coding RNA pair with each other. Pairing means complementary bases: A with U, or C with G; G–U wobble is excluded. Each base can have at most one partner, pairs cannot cross, and paired positions must satisfy the supplied minimum-loop value. Two stretches of the same RNA align in opposite directions. In an uninterrupted stem, the second stretch is the reverse complement of the first when both are written 5′→3′. For example, `5′-GGA-3′` aligns with `3′-CCU-5′`; writing that partner 5′→3′ gives `UCC`. Bulges or other unpaired bases can interrupt this relationship. The minimum-loop value is the number of positions that must lie between the endpoints of any pair. Search for the largest possible number of pairs under these rules and return a dot-bracket representation. This is a small algorithmic folding model, not a claim about the RNA’s physical structure. Exercise scope This follows the historical Nussinov approach to RNA folding: choose compatible, noncrossing base pairs by combining solutions for shorter sequence segments. The backbone geometry motivates restrictions on very short loops. Maximizing pair count is a deliberate simplification; thermodynamic methods additionally account for stacking and loop energies. Method references: Nussinov and colleagues: Algorithms for Loop Matchings, ViennaRNA. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `min_loop`: `3`. Procedure. Use interval dynamic programming: either leave a base unpaired or pair it with each permitted complementary partner, combining the best inside and remaining intervals. Maximize pair count, then choose the first full annotation in the stated character order. Final answer: (.(((.((((((((...)...)))..))).((...)))))).).. Checking the answer. The structure must have the maximum allowed pair count and be the first optimal dot-bracket string under the declared ASCII order. Easy example · references supplied Prompt ## Task Find a noncrossing RNA structure with the largest possible number of canonical Watson–Crick pairs. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "min_loop": 3 } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Choose canonical Watson–Crick pairs with no nucleotide paired more than once. - Pairs may be nested but must not cross: pairs `(i,j)` and `(k,l)` cannot coexist when `i < k < j < l`. - Each selected pair must have at least `min_loop` sequence positions between its endpoints: `j - i - 1 >= min_loop`. - Maximize the number of selected pairs, then apply the annotation tie-break below. ## Selection Maximize the allowed pair count first. Among optimal structures, choose the ASCII-lexicographically smallest full dot-bracket string, comparing characters in the order opening parenthesis, closing parenthesis, then dot. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ### Dot-bracket notation Matching parentheses mark paired RNA positions; a dot marks an unpaired position. Each closing parenthesis pairs with the most recent unmatched opening parenthesis. ## Answer - Return one continuous dot-bracket string with one character per RNA base. Use parentheses for the selected noncrossing pairs and dots for unpaired bases. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: (.(((.((((((((...)...)))..))).((...)))))).)..` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The answer optimizes the stated complementary-pair count; it is not a claim about the in vivo trpL RNA fold. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find a noncrossing RNA structure with the largest possible number of canonical Watson–Crick pairs. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "min_loop": {{min_loop}} } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Choose canonical Watson–Crick pairs with no nucleotide paired more than once. - Pairs may be nested but must not cross: pairs `(i,j)` and `(k,l)` cannot coexist when `i < k < j < l`. - Each selected pair must have at least `min_loop` sequence positions between its endpoints: `j - i - 1 >= min_loop`. - Maximize the number of selected pairs, then apply the annotation tie-break below. ## Selection Maximize the allowed pair count first. Among optimal structures, choose the ASCII-lexicographically smallest full dot-bracket string, comparing characters in the order opening parenthesis, closing parenthesis, then dot. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ### Dot-bracket notation Matching parentheses mark paired RNA positions; a dot marks an unpaired position. Each closing parenthesis pairs with the most recent unmatched opening parenthesis. ## Answer - Return one continuous dot-bracket string with one character per RNA base. Use parentheses for the selected noncrossing pairs and dots for unpaired bases. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Find a noncrossing RNA structure with the largest possible number of canonical Watson–Crick pairs. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "min_loop": 3 } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Choose canonical Watson–Crick pairs with no nucleotide paired more than once. - Pairs may be nested but must not cross: pairs `(i,j)` and `(k,l)` cannot coexist when `i < k < j < l`. - Each selected pair must have at least `min_loop` sequence positions between its endpoints: `j - i - 1 >= min_loop`. - Maximize the number of selected pairs, then apply the annotation tie-break below. ## Selection Maximize the allowed pair count first. Among optimal structures, choose the ASCII-lexicographically smallest full dot-bracket string, comparing characters in the order opening parenthesis, closing parenthesis, then dot. ## Answer - Return one continuous dot-bracket string with one character per RNA base. Use parentheses for the selected noncrossing pairs and dots for unpaired bases. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: (.(((.((((((((...)...)))..))).((...)))))).)..` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The answer optimizes the stated complementary-pair count; it is not a claim about the in vivo trpL RNA fold. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find a noncrossing RNA structure with the largest possible number of canonical Watson–Crick pairs. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "min_loop": {{min_loop}} } ``` ## Rules - Transcribe the CDS to obtain the coding RNA used in this task. - Choose canonical Watson–Crick pairs with no nucleotide paired more than once. - Pairs may be nested but must not cross: pairs `(i,j)` and `(k,l)` cannot coexist when `i < k < j < l`. - Each selected pair must have at least `min_loop` sequence positions between its endpoints: `j - i - 1 >= min_loop`. - Maximize the number of selected pairs, then apply the annotation tie-break below. ## Selection Maximize the allowed pair count first. Among optimal structures, choose the ASCII-lexicographically smallest full dot-bracket string, comparing characters in the order opening parenthesis, closing parenthesis, then dot. ## Answer - Return one continuous dot-bracket string with one character per RNA base. Use parentheses for the selected noncrossing pairs and dots for unpaired bases. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous dot-bracket string with one character per RNA base. Use parentheses for the selected noncrossing pairs and dots for unpaired bases. Selection Maximize the allowed pair count first. Among optimal structures, choose the ASCII-lexicographically smallest full dot-bracket string, comparing characters in the order opening parenthesis, closing parenthesis, then dot. Formal fields and scoring Inputs `rna` coding RNA `min_loop` minimum unpaired positions enclosed by each pair Rules & boundaries nussinov_wc_v1: maximize number of canonical pairs; one partner/base; noncrossing pairs; every pair j−i−1≥min_loop. No pseudoknots; use the declared dot-bracket tie-break. Maximize the allowed pair count first. Among optimal structures, choose the ASCII-lexicographically smallest full dot-bracket string, comparing characters in the order opening parenthesis, closing parenthesis, then dot. Design objective maximize canonical pair count Scoring `exact_match` · complete parsed answer Reference blocks Coding DNA to RNA, Canonical RNA base pairs, Dot-bracket notation Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "rna.max_pair_fold", "output_format": "scalar", "scorer": "exact_match", "family": "rna", "domain": "rna", "objective": "maximize canonical pair count", "selection_rule": "Maximize the allowed pair count first. Among optimal structures, choose the ASCII-lexicographically smallest full dot-bracket string, comparing characters in the order opening parenthesis, closing parenthesis, then dot." }

06 /

Synonymous redesign

Synonymous design across DNA and RNA constraints.

15 tasks

redesign task definitions
Task	Description	Expand details
Rewrite with a specified number of DNA changes`design.synonymous_distance`	Rewrite a CDS to encode the same protein while changing a requested number of DNA positions.
Rewrite with a specified number of DNA changes The question supplies a CDS and a permitted range for the number of changed bases. This range is the edit budget: it counts DNA positions whose letters differ between the original and rewritten sequences, not the number of codons touched. Use synonymous codons to satisfy that range while keeping the complete protein unchanged. The first and terminal stop codons stay fixed, and no bases may be inserted or deleted. This isolates how much DNA sequence can change without changing the encoded product. Exercise scope An exact number of DNA changes has no inherent biological benefit. The useful component is enforcing an edit budget while preserving a protein: joint RNA-and-codon optimization uses that constraint while pursuing a separate sequence-design objective. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `edit_bounds`: `[1, 2]`. Procedure. Enumerate the 1,152 synonymous minigene encodings with fixed terminal codons. Keep only sequences differing from the input at one or two bases, then choose the first whole DNA string in A<C<G<T order. Final answer: ATGAAAGCAATATTCGTACTATGA Checking the answer. The changed-position count must lie within the supplied inclusive range, with protein and fixed codons preserved. Select the first feasible DNA string under the stated order; this need not use the fewest edits. Easy example · references supplied Prompt ## Task Rewrite the CDS synonymously while keeping the number of changed DNA bases within `edit_bounds`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "edit_bounds": [1, 2] } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Count the nucleotide positions where the rewritten CDS differs from the original. This count must lie between both inclusive `edit_bounds`. - Use substitutions only, with no insertions or deletions. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCAATATTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously while keeping the number of changed DNA bases within `edit_bounds`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "edit_bounds": {{edit_bounds}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Count the nucleotide positions where the rewritten CDS differs from the original. This count must lie between both inclusive `edit_bounds`. - Use substitutions only, with no insertions or deletions. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Rewrite the CDS synonymously while keeping the number of changed DNA bases within `edit_bounds`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "edit_bounds": [1, 2] } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Count the nucleotide positions where the rewritten CDS differs from the original. This count must lie between both inclusive `edit_bounds`. - Use substitutions only, with no insertions or deletions. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCAATATTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously while keeping the number of changed DNA bases within `edit_bounds`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "edit_bounds": {{edit_bounds}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Count the nucleotide positions where the rewritten CDS differs from the original. This count must lie between both inclusive `edit_bounds`. - Use substitutions only, with no insertions or deletions. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `edit_bounds` inclusive [min,max] Hamming distance from original Rules & boundaries Same length; unchanged first/last codons; protein identical; Hamming distance within bounds. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.synonymous_distance", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "multi", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." } Used in: Balance an RNA structural constraint and codon preference
Rewrite using permitted codons`design.codon_allowlist`	Replace internal codons with synonymous alternatives drawn only from a supplied permitted list.
Rewrite using permitted codons The question supplies a list of DNA codons that may be used at internal protein-coding positions. This can model removing selected codons from a construct while leaving the amino-acid sequence intact. For each internal residue, choose an allowed codon that encodes the same amino acid. Keep the first and terminal stop codons unchanged. If a required amino acid has no permitted encoding at a necessary position, report that the requested rewrite is impossible. Practical context DNA Chisel can exclude codons that are rare in a chosen host. That provides a concrete source for an allowlist: retain the other synonymous codons while preserving the protein. The host reference or design policy must be supplied; an excluded codon is not intrinsically undesirable in every organism. Method reference: DNA Chisel specifications. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `allowed_codons`: supplied table shown in the prompt. Procedure. At each of the six internal positions, intersect the supplied allowed codons with synonymous choices. Compare the complete permitted DNA strings and choose the first in the stated order. Final answer: ATGAAGGCCATTTTCGTGCTGTGA Checking the answer. Every internal codon must be allowed, and the complete protein must remain identical. The selected sequence must also satisfy the declared tie-break. Easy example · references supplied Prompt ## Task Rewrite the CDS synonymously using the allowed codons at each internal position. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. Numeric keys in position maps are 1-based positions in the complete protein or its aligned codon sequence, including the initiator as position 1. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "allowed_codons": { "2": ["AAG"], "3": ["GCC", "GCT"], "4": ["ATT"], "5": ["TTC"], "6": ["GTG"], "7": ["CTG"] } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only internal sense codons may change. For each such position, choose a synonymous codon from its `allowed_codons` entry. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAGGCCATTTTCGTGCTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously using the allowed codons at each internal position. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. Numeric keys in position maps are 1-based positions in the complete protein or its aligned codon sequence, including the initiator as position 1. ```json { "cds": "{{cds}}", "allowed_codons": {{allowed_codons}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only internal sense codons may change. For each such position, choose a synonymous codon from its `allowed_codons` entry. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Rewrite the CDS synonymously using the allowed codons at each internal position. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. Numeric keys in position maps are 1-based positions in the complete protein or its aligned codon sequence, including the initiator as position 1. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "allowed_codons": { "2": ["AAG"], "3": ["GCC", "GCT"], "4": ["ATT"], "5": ["TTC"], "6": ["GTG"], "7": ["CTG"] } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only internal sense codons may change. For each such position, choose a synonymous codon from its `allowed_codons` entry. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAGGCCATTTTCGTGCTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously using the allowed codons at each internal position. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. Numeric keys in position maps are 1-based positions in the complete protein or its aligned codon sequence, including the initiator as position 1. ```json { "cds": "{{cds}}", "allowed_codons": {{allowed_codons}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only internal sense codons may change. For each such position, choose a synonymous codon from its `allowed_codons` entry. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `allowed_codons` allowed DNA codons for internal positions Rules & boundaries Internal sense codons must be allowed. First/last fixed. Feasibility determined exactly; select the first feasible encoding in DNA order. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.codon_allowlist", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "dna", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." }
Optimize supplied codon preferences`design.codon_preference`	Choose synonymous internal codons to maximize a supplied codon-preference score while retaining the protein.
Optimize supplied codon preferences The question provides positive numerical weights for codons. These are the preference system for this particular instance; there is no need to infer a host’s preferences or predict expression. For each internal sense position, choose a synonymous codon to maximize the defined codon adaptation index. The first and terminal stop codons are fixed. This simple version has no additional structure or motif constraints; those enter explicitly in compositions. Practical context This is the design counterpart of codon adaptation measurement: choose synonymous codons according to a host-derived preference reference when preparing a gene for expression. The joint-design composition adds an RNA constraint; methods such as LinearDesign extend that idea using a thermodynamic folding objective. Method reference: LinearDesign. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `weights`: supplied table shown in the prompt. Procedure. Evaluate the product of the six internal codon weights for each synonymous encoding. Maximizing that product also maximizes its sixth-root CAI; break any tie by the complete DNA string. Final answer: ATGAAAGCGATTTTCGTGCTGTGA Checking the answer. The candidate must encode the original protein and achieve the highest possible score. The selected sequence must also satisfy the declared tie-break. Easy example · references supplied Prompt ## Task Maximize supplied-weight CAI by synonymous recoding. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "weights": { "AAA": 1.0, "AAG": 0.3, "GCA": 0.2, "GCC": 0.3, "GCG": 1.0, "GCT": 0.5, "ATA": 0.2, "ATC": 0.3, "ATT": 1.0, "TTC": 1.0, "TTT": 0.3, "GTA": 0.2, "GTC": 0.3, "GTG": 1.0, "GTT": 0.5, "CTA": 0.2, "CTC": 0.3, "CTG": 1.0, "CTT": 0.5, "TTA": 0.6, "TTG": 0.7 } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Maximize CAI over internal sense codons using `weights`. Exclude the initiator and terminal stop from the objective. ## Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Codon adaptation index For n included codons with supplied positive weights w₁,…,wₙ, CAI = (w₁ × … × wₙ)^(1/n), the geometric mean of those weights. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCGATTTTCGTGCTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Weights are supplied illustrative preferences, not measured host expression data. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Maximize supplied-weight CAI by synonymous recoding. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "weights": {{weights}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Maximize CAI over internal sense codons using `weights`. Exclude the initiator and terminal stop from the objective. ## Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Codon adaptation index For n included codons with supplied positive weights w₁,…,wₙ, CAI = (w₁ × … × wₙ)^(1/n), the geometric mean of those weights. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Maximize supplied-weight CAI by synonymous recoding. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "weights": { "AAA": 1.0, "AAG": 0.3, "GCA": 0.2, "GCC": 0.3, "GCG": 1.0, "GCT": 0.5, "ATA": 0.2, "ATC": 0.3, "ATT": 1.0, "TTC": 1.0, "TTT": 0.3, "GTA": 0.2, "GTC": 0.3, "GTG": 1.0, "GTT": 0.5, "CTA": 0.2, "CTC": 0.3, "CTG": 1.0, "CTT": 0.5, "TTA": 0.6, "TTG": 0.7 } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Maximize CAI over internal sense codons using `weights`. Exclude the initiator and terminal stop from the objective. ## Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCGATTTTCGTGCTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Weights are supplied illustrative preferences, not measured host expression data. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Maximize supplied-weight CAI by synonymous recoding. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "weights": {{weights}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Maximize CAI over internal sense codons using `weights`. Exclude the initiator and terminal stop from the objective. ## Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `weights` positive weight for every sense codon Rules & boundaries First/last fixed; maximize CAI across internal sense codons. Break equal objective values by the declared DNA order. After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. Design objective maximize supplied-weight CAI Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons, Codon adaptation index Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.codon_preference", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "dna", "objective": "maximize supplied-weight CAI", "selection_rule": "After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding." }
Match positional codon-preference ranges`design.codon_profile`	Rewrite internal codons so each position meets its own supplied codon-preference range, while preserving the protein.
Match positional codon-preference ranges A single whole-CDS average does not specify where high- or low-preference codons occur. This task instead supplies preference weights and a separate acceptable weight interval for each internal codon position. Choose a synonymous codon whose weight lies inside that position’s interval. A request can deliberately retain lower-preference positions rather than optimizing every position upward. The ranges are supplied design targets, not something the model must infer from a natural expression pattern. Practical context Codon harmonization methods, such as CHARMING, try to reproduce a local pattern of codon usage when adapting a gene to another host, rather than maximizing every codon’s preference. The supplied positional ranges make that design idea testable here; they are a codon-usage proxy, not measured translation speeds. Method reference: CHARMING. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `weights`: supplied table shown in the prompt; `position_bounds`: supplied table shown in the prompt. Procedure. For each internal position, retain synonymous codons whose individual weight is inside that position’s supplied range. Combine the choices and select the first complete DNA string; do not average the ranges. Final answer: ATGAAGGCGATATTCGTACTGTGA Checking the answer. Every positional interval must be satisfied, with the protein and terminal codons preserved. Report IMPOSSIBLE if no assignment exists. Easy example · references supplied Prompt ## Task Rewrite the CDS synonymously so each internal codon’s preference weight falls within its assigned range. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. Numeric keys in position maps are 1-based positions in the complete protein or its aligned codon sequence, including the initiator as position 1. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "weights": { "AAA": 1.0, "AAG": 0.3, "GCA": 0.2, "GCC": 0.3, "GCG": 1.0, "GCT": 0.5, "ATA": 0.2, "ATC": 0.3, "ATT": 1.0, "TTC": 1.0, "TTT": 0.3, "GTA": 0.2, "GTC": 0.3, "GTG": 1.0, "GTT": 0.5, "CTA": 0.2, "CTC": 0.3, "CTG": 1.0, "CTT": 0.5, "TTA": 0.6, "TTG": 0.7 }, "position_bounds": { "2": [0.2, 0.7], "3": [0.8, 1.0], "4": [0.2, 0.7], "5": [0.8, 1.0], "6": [0.2, 0.7], "7": [0.8, 1.0] } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - For each internal codon position, look up the chosen synonymous codon’s weight. It must lie within that position’s inclusive `position_bounds`. - Check each position separately; an average across positions does not satisfy the requirement. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAGGCGATATTCGTACTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Weights and positional ranges are supplied illustrative inputs. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously so each internal codon’s preference weight falls within its assigned range. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. Numeric keys in position maps are 1-based positions in the complete protein or its aligned codon sequence, including the initiator as position 1. ```json { "cds": "{{cds}}", "weights": {{weights}}, "position_bounds": {{position_bounds}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - For each internal codon position, look up the chosen synonymous codon’s weight. It must lie within that position’s inclusive `position_bounds`. - Check each position separately; an average across positions does not satisfy the requirement. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Rewrite the CDS synonymously so each internal codon’s preference weight falls within its assigned range. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. Numeric keys in position maps are 1-based positions in the complete protein or its aligned codon sequence, including the initiator as position 1. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "weights": { "AAA": 1.0, "AAG": 0.3, "GCA": 0.2, "GCC": 0.3, "GCG": 1.0, "GCT": 0.5, "ATA": 0.2, "ATC": 0.3, "ATT": 1.0, "TTC": 1.0, "TTT": 0.3, "GTA": 0.2, "GTC": 0.3, "GTG": 1.0, "GTT": 0.5, "CTA": 0.2, "CTC": 0.3, "CTG": 1.0, "CTT": 0.5, "TTA": 0.6, "TTG": 0.7 }, "position_bounds": { "2": [0.2, 0.7], "3": [0.8, 1.0], "4": [0.2, 0.7], "5": [0.8, 1.0], "6": [0.2, 0.7], "7": [0.8, 1.0] } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - For each internal codon position, look up the chosen synonymous codon’s weight. It must lie within that position’s inclusive `position_bounds`. - Check each position separately; an average across positions does not satisfy the requirement. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAGGCGATATTCGTACTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Weights and positional ranges are supplied illustrative inputs. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously so each internal codon’s preference weight falls within its assigned range. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. Numeric keys in position maps are 1-based positions in the complete protein or its aligned codon sequence, including the initiator as position 1. ```json { "cds": "{{cds}}", "weights": {{weights}}, "position_bounds": {{position_bounds}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - For each internal codon position, look up the chosen synonymous codon’s weight. It must lie within that position’s inclusive `position_bounds`. - Check each position separately; an average across positions does not satisfy the requirement. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `weights` supplied per-codon preference weights `position_bounds` inclusive weight bounds per internal sense-codon index Rules & boundaries Every chosen internal codon weight must satisfy its positional bounds; first/last fixed. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.codon_profile", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "dna", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." }
Control local GC by recoding`design.gc_windows`	Synonymously rewrite a CDS so every local DNA window has GC content within a supplied range.
Control local GC by recoding The question gives a window length and a minimum and maximum GC fraction. These define local composition requirements rather than a single whole-gene target. Change internal codons without changing the protein, then check every full window one base at a time across the final DNA. Overlapping windows couple the choices: fixing one region can create a problem in its neighbor. The first and terminal stop codons remain fixed. Practical context Synthetic-gene design often needs to reduce local GC extremes while keeping the encoded product fixed. The cloning-and-GC composition combines this requirement with restriction-site removal, reflecting the simultaneous sequence constraints handled by practical construct-design tools. Method references: Twist gene-design resources, DNA Chisel examples. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `window`: `9`; `gc_bounds`: `["3/9", "4/9"]`. Procedure. The original minigene violates the stricter GC bounds. For each synonymous candidate, check every one of its 16 nine-base windows against the inclusive 3/9–4/9 range. Choose the first DNA string that passes every window. Final answer: ATGAAAGCAATCTTCGTACTATGA Checking the answer. The same final CDS must pass every window and preserve the original protein. A good average GC value alone is insufficient. Easy example · references supplied Prompt ## Task Rewrite the CDS synonymously so every full-length window meets the supplied GC bounds. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. GC bounds are inclusive fractions. A string such as `"1/3"` denotes that exact rational value. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "window": 9, "gc_bounds": ["3/9", "4/9"] } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Move a window of length `window` along the entire final CDS one base at a time, checking every window that fits completely. - Every window must meet the inclusive GC-fraction bounds, including windows that overlap the fixed first codon or terminal stop. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### GC fraction GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCAATCTTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously so every full-length window meets the supplied GC bounds. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. GC bounds are inclusive fractions. A string such as `"1/3"` denotes that exact rational value. ```json { "cds": "{{cds}}", "window": {{window}}, "gc_bounds": {{gc_bounds}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Move a window of length `window` along the entire final CDS one base at a time, checking every window that fits completely. - Every window must meet the inclusive GC-fraction bounds, including windows that overlap the fixed first codon or terminal stop. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### GC fraction GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Rewrite the CDS synonymously so every full-length window meets the supplied GC bounds. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. GC bounds are inclusive fractions. A string such as `"1/3"` denotes that exact rational value. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "window": 9, "gc_bounds": ["3/9", "4/9"] } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Move a window of length `window` along the entire final CDS one base at a time, checking every window that fits completely. - Every window must meet the inclusive GC-fraction bounds, including windows that overlap the fixed first codon or terminal stop. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCAATCTTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously so every full-length window meets the supplied GC bounds. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. GC bounds are inclusive fractions. A string such as `"1/3"` denotes that exact rational value. ```json { "cds": "{{cds}}", "window": {{window}}, "gc_bounds": {{gc_bounds}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Move a window of length `window` along the entire final CDS one base at a time, checking every window that fits completely. - Every window must meet the inclusive GC-fraction bounds, including windows that overlap the fixed first codon or terminal stop. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `window` width `gc_bounds` inclusive fraction bounds Rules & boundaries All full stride-1 windows obey bounds, including windows overlapping fixed terminal codons. Small instances have certified feasibility. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons, GC fraction Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.gc_windows", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "dna", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." } Used in: Remove cloning sites while controlling GC
Remove specified DNA motifs`design.remove_dna_motifs`	Remove every occurrence of specified DNA patterns using synonymous changes, preserving the encoded protein.
Remove specified DNA motifs A construct may contain a recognition site for an enzyme that will be used during cloning. For example, EcoRI recognizes GAATTC, the site in the worked example. The question supplies the unwanted patterns and whether to search the coding strand alone or both strands. Change internal codons synonymously until every forbidden occurrence is absent from the final CDS. Search across codon boundaries as well as within codons, and check that the edits did not create a new copy elsewhere. Preserve the first and terminal stop codons. Practical context Golden Gate assembly requires removing unwanted internal recognition sites for the Type IIS enzyme being used; otherwise the insert can be cut internally. This synonymous cleanup is often called domestication. The combined design task also controls GC while removing the sites. Method reference: GoldenMutagenesis domestication. Task Outline Inputs. `cds`: `ATG AAA GAA TTC TTC GTA CTG TGA`; `forbidden_patterns`: `["GAATTC"]`; `strand_mode`: `both`. Procedure. Search each synonymous candidate for GAATTC on both strands, including matches across codon boundaries. Discard any candidate containing the EcoRI recognition site and select the first surviving full DNA string. Final answer: ATGAAAGAATTTTTCGTACTATGA Checking the answer. The candidate must preserve the protein and contain no forbidden hits under the requested strand rule. The selected sequence must also satisfy the declared tie-break. Easy example · references supplied Prompt ## Task Remove every forbidden DNA pattern through synonymous changes to the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GAA TTC TTC GTA CTG TGA", "forbidden_patterns": ["GAATTC"], "strand_mode": "both" } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Remove every occurrence of every `forbidden_patterns` entry, including overlapping matches and matches that span codon boundaries. - For `strand_mode=both`, also remove occurrences of each pattern’s reverse complement. For `forward`, search only the supplied coding strand. - Check the complete final DNA sequence after all changes. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGAATTTTTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The minigene has A3E and I4F substitutions (GCA→GAA, ATT→TTC), creating the real EcoRI recognition sequence GAATTC at bases 7–12. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Remove every forbidden DNA pattern through synonymous changes to the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "forbidden_patterns": {{forbidden_patterns}}, "strand_mode": "{{strand_mode}}" } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Remove every occurrence of every `forbidden_patterns` entry, including overlapping matches and matches that span codon boundaries. - For `strand_mode=both`, also remove occurrences of each pattern’s reverse complement. For `forward`, search only the supplied coding strand. - Check the complete final DNA sequence after all changes. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Remove every forbidden DNA pattern through synonymous changes to the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GAA TTC TTC GTA CTG TGA", "forbidden_patterns": ["GAATTC"], "strand_mode": "both" } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Remove every occurrence of every `forbidden_patterns` entry, including overlapping matches and matches that span codon boundaries. - For `strand_mode=both`, also remove occurrences of each pattern’s reverse complement. For `forward`, search only the supplied coding strand. - Check the complete final DNA sequence after all changes. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGAATTTTTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The minigene has A3E and I4F substitutions (GCA→GAA, ATT→TTC), creating the real EcoRI recognition sequence GAATTC at bases 7–12. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Remove every forbidden DNA pattern through synonymous changes to the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "forbidden_patterns": {{forbidden_patterns}}, "strand_mode": "{{strand_mode}}" } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Remove every occurrence of every `forbidden_patterns` entry, including overlapping matches and matches that span codon boundaries. - For `strand_mode=both`, also remove occurrences of each pattern’s reverse complement. For `forward`, search only the supplied coding strand. - Check the complete final DNA sequence after all changes. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `forbidden_patterns` IUPAC DNA patterns `strand_mode` forward \| both Rules & boundaries Remove every hit of every pattern, including cross-codon hits and reverse complements if both. Terminal codons fixed. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons, IUPAC DNA base sets, DNA reverse complement Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.remove_dna_motifs", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "dna", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." } Used in: Remove cloning sites while controlling GC Encode a protein change and remove cloning sites
Introduce a DNA recognition site`design.insert_dna_motif`	Create a specified DNA recognition sequence at a requested CDS position through synonymous changes.
Introduce a DNA recognition site A restriction-enzyme recognition site can be introduced as a marker for screening an engineered sequence. The worked example uses EcoRI’s real GAATTC site. The input specifies the literal recognition sequence and the DNA position where it must begin. Choose synonymous internal codons so the required bases appear at the requested location, even if the site spans several codons. Other copies of the site are allowed in this version. Keep the first and terminal stop codons fixed. Practical context Introducing a synonymous restriction site can create a convenient way to screen an engineered allele by restriction digestion. This task isolates site creation without changing the protein. Motif scanning can then confirm where the recognition sequence occurs. Method reference: Restriction-site markers in engineered alleles. Task Outline Inputs. `cds`: `ATG AAA GAG TTC TTC GTA CTG TGA`; `pattern`: `GAATTC`; `start`: `7`. Procedure. Require GAATTC exactly at DNA bases 7–12 while preserving the protein. This fixes codons 3–4 to GAA TTC; choose the first complete synonymous encoding satisfying that site. Final answer: ATGAAAGAATTCTTCGTACTATGA Checking the answer. The required site must appear at the specified start and all protein-preservation conditions must hold. Otherwise return IMPOSSIBLE. Easy example · references supplied Prompt ## Task Introduce the specified DNA pattern at position `start` through synonymous changes to the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GAG TTC TTC GTA CTG TGA", "pattern": "GAATTC", "start": 7 } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - The literal DNA string `pattern` must begin at the supplied 1-based position `start` on the forward coding strand. - Other copies of that pattern elsewhere in the final CDS are allowed. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGAATTCTTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Use the same A3E/I4F cloning substrate, with synonymous GAA→GAG at codon 3 removing the EcoRI site before the task. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Introduce the specified DNA pattern at position `start` through synonymous changes to the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pattern": "{{pattern}}", "start": {{start}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - The literal DNA string `pattern` must begin at the supplied 1-based position `start` on the forward coding strand. - Other copies of that pattern elsewhere in the final CDS are allowed. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Introduce the specified DNA pattern at position `start` through synonymous changes to the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GAG TTC TTC GTA CTG TGA", "pattern": "GAATTC", "start": 7 } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - The literal DNA string `pattern` must begin at the supplied 1-based position `start` on the forward coding strand. - Other copies of that pattern elsewhere in the final CDS are allowed. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGAATTCTTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Use the same A3E/I4F cloning substrate, with synonymous GAA→GAG at codon 3 removing the EcoRI site before the task. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Introduce the specified DNA pattern at position `start` through synonymous changes to the CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pattern": "{{pattern}}", "start": {{start}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - The literal DNA string `pattern` must begin at the supplied 1-based position `start` on the forward coding strand. - Other copies of that pattern elsewhere in the final CDS are allowed. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `pattern` literal DNA sequence `start` required 1-based forward-strand start Rules & boundaries Literal pattern must occur at start. Other occurrences are allowed. First/last codons fixed. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.insert_dna_motif", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "dna", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." }
Break homopolymers and repeats`design.break_repeats`	Use synonymous substitutions to remove overly long runs of one base and excessive repeated DNA substrings.
Break homopolymers and repeats The question gives the longest permitted run of identical bases and limits on how many times a substring of a given length may occur. These are sequence-design requirements supplied for the instance. Recode internal codons without changing the protein. Count runs and repeated substrings across the entire final CDS, including codon junctions and overlapping occurrences. The first and terminal stop codons cannot be changed. Practical context Long homopolymers and repeated sequences can make DNA synthesis and amplification more difficult. Synonymous recoding can reduce that sequence complexity without changing the protein. The benchmark’s explicit repeat rules represent one part of the broader manufacturability screening used by synthesis providers. Method reference: Twist gene-design resources. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `max_run`: `2`; `max_kmer_copies`: `{"3": 2}`. Procedure. For each synonymous minigene, check the longest identical-base run and count all overlapping three-base substrings. Reject runs longer than two and any triplet appearing more than twice, then choose the first valid DNA string. Final answer: ATGAAGGCAATATTCGTACTATGA Checking the answer. Every repeat limit must hold on the final sequence while translation stays unchanged. Satisfying only the homopolymer limit is not enough when substring limits are also supplied. Easy example · references supplied Prompt ## Task Rewrite the CDS synonymously to satisfy the limits on runs of identical bases and repeated DNA substrings. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "max_run": 2, "max_kmer_copies": { "3": 2 } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - No run of consecutive identical DNA bases may be longer than `max_run`. - For each length k listed in `max_kmer_copies`, count every occurrence of each distinct k-base substring across the full CDS, including overlaps. No substring may exceed the corresponding occurrence limit. - Count a substring and its reverse complement separately. Include runs and repeats that cross codon boundaries. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAGGCAATATTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously to satisfy the limits on runs of identical bases and repeated DNA substrings. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "max_run": {{max_run}}, "max_kmer_copies": {{max_kmer_copies}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - No run of consecutive identical DNA bases may be longer than `max_run`. - For each length k listed in `max_kmer_copies`, count every occurrence of each distinct k-base substring across the full CDS, including overlaps. No substring may exceed the corresponding occurrence limit. - Count a substring and its reverse complement separately. Include runs and repeats that cross codon boundaries. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Rewrite the CDS synonymously to satisfy the limits on runs of identical bases and repeated DNA substrings. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "max_run": 2, "max_kmer_copies": { "3": 2 } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - No run of consecutive identical DNA bases may be longer than `max_run`. - For each length k listed in `max_kmer_copies`, count every occurrence of each distinct k-base substring across the full CDS, including overlaps. No substring may exceed the corresponding occurrence limit. - Count a substring and its reverse complement separately. Include runs and repeats that cross codon boundaries. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAGGCAATATTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously to satisfy the limits on runs of identical bases and repeated DNA substrings. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "max_run": {{max_run}}, "max_kmer_copies": {{max_kmer_copies}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - No run of consecutive identical DNA bases may be longer than `max_run`. - For each length k listed in `max_kmer_copies`, count every occurrence of each distinct k-base substring across the full CDS, including overlaps. No substring may exceed the corresponding occurrence limit. - Count a substring and its reverse complement separately. Include runs and repeats that cross codon boundaries. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `max_run` maximum identical-base run `max_kmer_copies` map: k to maximum occurrences of any k-mer Rules & boundaries Count overlapping k-mer occurrences; check entire CDS; no reverse-complement merging. First/last fixed. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.break_repeats", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "dna", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." }
Control coding-RNA motifs`design.rna_motif_control`	Synonymously rewrite DNA so its coding RNA contains specified numbers of supplied RNA patterns.
Control coding-RNA motifs The patterns are RNA strings using U, and each comes with an acceptable minimum and maximum number of occurrences. This can request removing a pattern, introducing one, or controlling its copy count. Choose synonymous DNA codons, transcribe the candidate, and count each RNA pattern in that final RNA. Include overlapping matches and matches crossing codon boundaries. Keep the protein and first and terminal stop codons unchanged. Exercise scope If a supplied pattern is a probe’s exact complementary target, synonymous recoding can remove that candidate binding sequence while preserving the protein. Guide-escape design extends this idea to near-matches. Dinucleotide control provides another concrete composition target; a generic unnamed pattern remains an abstract constraint. Method reference: NUPACK documentation. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `rna_patterns`: `["AAA", "AAG"]`; `count_bounds`: `{"AAA": [0, 0], "AAG": [1, 1]}`. Procedure. Transcribe each synonymous candidate and count all overlapping AAA and AAG occurrences, including codon junctions. Require zero AAA and exactly one AAG, then apply the DNA-order tie-break. Final answer: ATGAAGGCAATATTCGTACTATGA Checking the answer. All pattern counts must fall within their supplied ranges on the same final RNA, while the encoded protein is preserved. Easy example · references supplied Prompt ## Task Rewrite the CDS synonymously so its coding RNA meets every supplied pattern-count bound. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "rna_patterns": ["AAA", "AAG"], "count_bounds": { "AAA": [0, 0], "AAG": [1, 1] } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Transcribe each candidate CDS and count exact forward matches for every supplied RNA pattern. Include overlapping occurrences and occurrences spanning codon boundaries. - Each count must meet that pattern’s inclusive `count_bounds`. All patterns must satisfy their bounds in the same final RNA. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAGGCAATATTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously so its coding RNA meets every supplied pattern-count bound. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "rna_patterns": {{rna_patterns}}, "count_bounds": {{count_bounds}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Transcribe each candidate CDS and count exact forward matches for every supplied RNA pattern. Include overlapping occurrences and occurrences spanning codon boundaries. - Each count must meet that pattern’s inclusive `count_bounds`. All patterns must satisfy their bounds in the same final RNA. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Rewrite the CDS synonymously so its coding RNA meets every supplied pattern-count bound. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "rna_patterns": ["AAA", "AAG"], "count_bounds": { "AAA": [0, 0], "AAG": [1, 1] } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Transcribe each candidate CDS and count exact forward matches for every supplied RNA pattern. Include overlapping occurrences and occurrences spanning codon boundaries. - Each count must meet that pattern’s inclusive `count_bounds`. All patterns must satisfy their bounds in the same final RNA. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAGGCAATATTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously so its coding RNA meets every supplied pattern-count bound. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "rna_patterns": {{rna_patterns}}, "count_bounds": {{count_bounds}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Transcribe each candidate CDS and count exact forward matches for every supplied RNA pattern. Include overlapping occurrences and occurrences spanning codon boundaries. - Each count must meet that pattern’s inclusive `count_bounds`. All patterns must satisfy their bounds in the same final RNA. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `rna_patterns` literal RNA patterns `count_bounds` inclusive occurrence bounds per pattern Rules & boundaries Transcribe candidate to RNA; count overlapping exact forward matches including codon junctions. First/last fixed. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons, Coding DNA to RNA Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.rna_motif_control", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "rna", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." }
Control RNA dinucleotide counts`design.dinucleotide_counts`	Synonymously rewrite a CDS to meet requested counts of adjacent RNA-base pairs such as CG or UA.
Control RNA dinucleotide counts A dinucleotide is simply two consecutive bases along a sequence, not a structural base pair. CpG means a C followed by G; UpA means U followed by A. The question gives acceptable counts for selected RNA dinucleotides. Change synonymous DNA codons and count adjacent pairs in the resulting coding RNA at every position. A pair spanning two codons counts just like a pair inside a codon. Keep the protein and first and terminal stop codons fixed. Practical context Dinucleotide composition can affect how cellular factors recognize RNA; for example, ZAP recognizes CpG-containing RNA. A controlled synonymous panel uses count changes to construct testable comparisons while preserving protein identity and total GC. The count is an experimental design variable, rather than a prediction of an effect’s size. Method reference: Origin and evolution of the zinc finger antiviral protein. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `dinucleotide_bounds`: `{"CG": [2, 2], "GC": [1, 1]}`. Procedure. The input has only one CG. Count adjacent bases in each candidate’s coding RNA and require exactly two CG occurrences and exactly one GC occurrence. Select the first complete synonymous DNA string passing both counts. Final answer: ATGAAAGCGATATTCGTACTATGA Checking the answer. Every requested dinucleotide count must lie within its stated range in the final coding RNA. Easy example · references supplied Prompt ## Task Rewrite the CDS synonymously so its coding RNA meets every supplied dinucleotide-count bound. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "dinucleotide_bounds": { "CG": [2, 2], "GC": [1, 1] } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Count every adjacent pair of bases in the candidate coding RNA, moving one base at a time. Include pairs spanning codon boundaries. - Each requested dinucleotide must meet its inclusive count bounds. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCGATATTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously so its coding RNA meets every supplied dinucleotide-count bound. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "dinucleotide_bounds": {{dinucleotide_bounds}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Count every adjacent pair of bases in the candidate coding RNA, moving one base at a time. Include pairs spanning codon boundaries. - Each requested dinucleotide must meet its inclusive count bounds. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Rewrite the CDS synonymously so its coding RNA meets every supplied dinucleotide-count bound. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "dinucleotide_bounds": { "CG": [2, 2], "GC": [1, 1] } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Count every adjacent pair of bases in the candidate coding RNA, moving one base at a time. Include pairs spanning codon boundaries. - Each requested dinucleotide must meet its inclusive count bounds. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCGATATTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously so its coding RNA meets every supplied dinucleotide-count bound. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "dinucleotide_bounds": {{dinucleotide_bounds}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Count every adjacent pair of bases in the candidate coding RNA, moving one base at a time. Include pairs spanning codon boundaries. - Each requested dinucleotide must meet its inclusive count bounds. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `dinucleotide_bounds` map: RNA dinucleotide to inclusive count bounds Rules & boundaries Count adjacent pairs at stride 1, including pairs spanning codons; CpG means CG, UpA means UA. First/last fixed. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons, Coding DNA to RNA Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.dinucleotide_counts", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "rna", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." }
Repair pairing synonymously`design.pairing_repair`	Use synonymous CDS changes to make specified positions in its coding RNA form allowed base pairs.
Repair pairing synonymously The question supplies the intended RNA pairs and the DNA positions that may be edited. A mismatch in the supplied map is known in advance; the model does not have to discover the structure or identify a biological compensatory mutation. Choose changes at the permitted positions so every listed RNA pair becomes canonical, while the CDS still encodes its original protein. The first and terminal stop codons remain fixed. A change that repairs one pair but changes an amino acid is not a valid repair here. Practical context This has a direct experimental precedent: a study of bacterial rpsT mRNA found that some fitness costs of synonymous mutations could be compensated by second synonymous changes restoring predicted base pairing. Repair and verification isolates the sequence-design logic: restore the proposed RNA contact without changing the encoded protein. Method reference: Synonymous mutations and restoration of mRNA base pairing. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `pairs`: `[[3, 9]]`; `editable_positions`: `[9]`. Procedure. RNA base 3 is G and base 9 is initially A. Change only DNA base 9: Ala codon GCA can become GCC without changing the protein, replacing the incompatible G–A contact with G–C. Final answer: ATGAAAGCCATTTTCGTACTGTGA Checking the answer. All specified pairs must be compatible, all DNA changes permitted, and the protein unchanged. The selected sequence must also satisfy the declared tie-break. Easy example · references supplied Prompt ## Task Use permitted synonymous DNA substitutions to make every listed RNA pair a canonical Watson–Crick pair. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "pairs": [ [3, 9] ], "editable_positions": [9] } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Change DNA bases only at `editable_positions`. - In the candidate coding RNA, every position pair listed in `pairs` must be a canonical Watson–Crick pair. Exclude wobble. - Check the supplied pairs only; no folding prediction is required. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCCATTTTCGTACTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The supplied pair (3,9) is initially G–A; the only permitted DNA position is the third base of Ala codon 3. The map is an illustrative repair target. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Use permitted synonymous DNA substitutions to make every listed RNA pair a canonical Watson–Crick pair. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pairs": {{pairs}}, "editable_positions": {{editable_positions}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Change DNA bases only at `editable_positions`. - In the candidate coding RNA, every position pair listed in `pairs` must be a canonical Watson–Crick pair. Exclude wobble. - Check the supplied pairs only; no folding prediction is required. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Use permitted synonymous DNA substitutions to make every listed RNA pair a canonical Watson–Crick pair. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "pairs": [ [3, 9] ], "editable_positions": [9] } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Change DNA bases only at `editable_positions`. - In the candidate coding RNA, every position pair listed in `pairs` must be a canonical Watson–Crick pair. Exclude wobble. - Check the supplied pairs only; no folding prediction is required. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCCATTTTCGTACTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The supplied pair (3,9) is initially G–A; the only permitted DNA position is the third base of Ala codon 3. The map is an illustrative repair target. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Use permitted synonymous DNA substitutions to make every listed RNA pair a canonical Watson–Crick pair. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pairs": {{pairs}}, "editable_positions": {{editable_positions}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Change DNA bases only at `editable_positions`. - In the candidate coding RNA, every position pair listed in `pairs` must be a canonical Watson–Crick pair. Exclude wobble. - Check the supplied pairs only; no folding prediction is required. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS, including any fixed disruptive mutation `pairs` required coding-RNA position pairs `editable_positions` allowed nucleotide positions Rules & boundaries All specified pairs canonical after T→U. Only editable_positions may change. First/last fixed; preserve protein of the provided CDS. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons, Coding DNA to RNA, Canonical RNA base pairs Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.pairing_repair", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "rna", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." } Used in: Repair RNA pairs and verify the protein
Optimize a fixed RNA pairing score`design.structure_score`	Choose synonymous codons to maximize or minimize a supplied score for a fixed coding-RNA pairing map.
Optimize a fixed RNA pairing score The RNA position pairs are fixed by the question. A numerical table assigns values to allowed pair types, and the question says whether to make their total larger or smaller. This is not a request to predict folding free energy. Search synonymous versions of the CDS, preserving the first and terminal stop codons. Every required pair must remain canonical. Among feasible candidates, choose one whose total pair score is optimal in the requested direction. Exercise scope This is an optimization exercise over a supplied map and artificial pair scores. The joint-design composition tests whether the solver can satisfy that objective alongside codon and edit constraints. Practical mRNA design uses richer folding objectives, as illustrated by LinearDesign. Method reference: LinearDesign. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `pairs`: `[[9, 18]]`; `pair_scores`: `{"AU": 2, "UA": 2, "CG": 3, "GC": 3}`; `direction`: `maximize`. Procedure. RNA positions 9 and 18 are the third bases of codons 3 and 6. Enumerate compatible synonymous choices, sum the supplied score for their fixed contact, maximize that score, and use the complete DNA string to break ties. Final answer: ATGAAAGCCATATTCGTGCTATGA Checking the answer. The final CDS must preserve the protein, satisfy every pair, and attain the optimum under the supplied score table. Easy example · references supplied Prompt ## Task Rewrite the CDS synonymously to optimize the score of the supplied RNA pairing map in the requested `direction`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "pairs": [ [9, 18] ], "pair_scores": { "AU": 2, "UA": 2, "CG": 3, "GC": 3 }, "direction": "maximize" } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Use the supplied pair map unchanged. Every listed pair must be canonical in the candidate coding RNA. - The score is the sum of the `pair_scores` entries for the bases at each listed pair. - For `direction=maximize`, choose the greatest total; for `minimize`, choose the least. ## Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCCATATTCGTGCTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Positions 9 and 18 are synonymous third bases. The fixed pair map and scores are supplied test inputs. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously to optimize the score of the supplied RNA pairing map in the requested `direction`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pairs": {{pairs}}, "pair_scores": {{pair_scores}}, "direction": "{{direction}}" } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Use the supplied pair map unchanged. Every listed pair must be canonical in the candidate coding RNA. - The score is the sum of the `pair_scores` entries for the bases at each listed pair. - For `direction=maximize`, choose the greatest total; for `minimize`, choose the least. ## Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Rewrite the CDS synonymously to optimize the score of the supplied RNA pairing map in the requested `direction`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "pairs": [ [9, 18] ], "pair_scores": { "AU": 2, "UA": 2, "CG": 3, "GC": 3 }, "direction": "maximize" } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Use the supplied pair map unchanged. Every listed pair must be canonical in the candidate coding RNA. - The score is the sum of the `pair_scores` entries for the bases at each listed pair. - For `direction=maximize`, choose the greatest total; for `minimize`, choose the least. ## Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCCATATTCGTGCTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Positions 9 and 18 are synonymous third bases. The fixed pair map and scores are supplied test inputs. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously to optimize the score of the supplied RNA pairing map in the requested `direction`. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pairs": {{pairs}}, "pair_scores": {{pair_scores}}, "direction": "{{direction}}" } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Use the supplied pair map unchanged. Every listed pair must be canonical in the candidate coding RNA. - The score is the sum of the `pair_scores` entries for the bases at each listed pair. - For `direction=maximize`, choose the greatest total; for `minimize`, choose the least. ## Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `pairs` required coding-RNA pair map `pair_scores` supplied canonical pair scores `direction` maximize \| minimize Rules & boundaries All pairs must be canonical. Optimize wc_additive_v1 total over fixed map; first/last fixed. Small exact-search instances only. Not physical stability. After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. Design objective direction over fixed-map additive pair score Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons, Coding DNA to RNA, Canonical RNA base pairs Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.structure_score", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "rna", "objective": "direction over fixed-map additive pair score", "selection_rule": "After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding." }
Disrupt a specified local hairpin`design.expose_rna_window`	Synonymously change a CDS to break specified RNA stem pairs that involve a selected region.
Disrupt a specified local hairpin The question supplies a candidate pairing map, an RNA interval of interest, and the DNA positions that may change. It asks for disruption of the listed stem pairs touching that region, rather than discovery of all possible alternative structures. Make permitted synonymous substitutions so every relevant listed pair becomes incompatible under the no-wobble pairing rule. Pairs outside the selected region need not be broken. Preserve the protein and fixed terminal codons. Exercise scope A complementary probe must compete with bases already paired inside its target RNA. Disrupting a proposed stem is therefore a first step toward designing an accessible binding region. Practical methods also evaluate alternative folds and binding energies; this task checks only the supplied contacts. Method reference: NUPACK documentation. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `candidate_pairs`: `[[4, 13], [5, 12], [6, 11]]`; `target_interval`: `[12, 12]`; `editable_positions`: `[12]`. Procedure. Of the three supplied A–U stem contacts, only (5,12) touches the target base 12. Change the Ile codon ATT to ATA at its third position: that contact becomes A–A while the encoded protein stays unchanged. Final answer: ATGAAAGCAATATTCGTACTGTGA Checking the answer. All targeted listed pairs must be broken without an unauthorized DNA or protein change. Passing does not establish physical accessibility if the RNA can adopt an unlisted fold. Easy example · references supplied Prompt ## Task Use permitted synonymous DNA substitutions to break the listed RNA stem pairs involving the target region. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "candidate_pairs": [ [4, 13], [5, 12], [6, 11] ], "target_interval": [12, 12], "editable_positions": [12] } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Treat `candidate_pairs` as a fixed checklist. Break every listed pair with at least one endpoint inside the inclusive RNA interval `target_interval`, so its two bases no longer form a canonical Watson–Crick pair. - Listed pairs with neither endpoint in the target interval do not have to be broken. - Do not predict a new RNA fold. Whether unlisted pairs could form does not affect the answer. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCAATATTCGTACTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The supplied three-pair stem initially consists of A–U contacts (4,13), (5,12), and (6,11). Expose base 12 by altering its synonymous codon position; this is not a prediction of the natural fold. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Use permitted synonymous DNA substitutions to break the listed RNA stem pairs involving the target region. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "candidate_pairs": {{candidate_pairs}}, "target_interval": {{target_interval}}, "editable_positions": {{editable_positions}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Treat `candidate_pairs` as a fixed checklist. Break every listed pair with at least one endpoint inside the inclusive RNA interval `target_interval`, so its two bases no longer form a canonical Watson–Crick pair. - Listed pairs with neither endpoint in the target interval do not have to be broken. - Do not predict a new RNA fold. Whether unlisted pairs could form does not affect the answer. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Use permitted synonymous DNA substitutions to break the listed RNA stem pairs involving the target region. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "candidate_pairs": [ [4, 13], [5, 12], [6, 11] ], "target_interval": [12, 12], "editable_positions": [12] } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Treat `candidate_pairs` as a fixed checklist. Break every listed pair with at least one endpoint inside the inclusive RNA interval `target_interval`, so its two bases no longer form a canonical Watson–Crick pair. - Listed pairs with neither endpoint in the target interval do not have to be broken. - Do not predict a new RNA fold. Whether unlisted pairs could form does not affect the answer. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCAATATTCGTACTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The supplied three-pair stem initially consists of A–U contacts (4,13), (5,12), and (6,11). Expose base 12 by altering its synonymous codon position; this is not a prediction of the natural fold. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Use permitted synonymous DNA substitutions to break the listed RNA stem pairs involving the target region. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "candidate_pairs": {{candidate_pairs}}, "target_interval": {{target_interval}}, "editable_positions": {{editable_positions}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Treat `candidate_pairs` as a fixed checklist. Break every listed pair with at least one endpoint inside the inclusive RNA interval `target_interval`, so its two bases no longer form a canonical Watson–Crick pair. - Listed pairs with neither endpoint in the target interval do not have to be broken. - Do not predict a new RNA fold. Whether unlisted pairs could form does not affect the answer. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `candidate_pairs` potential stem pairs in a supplied map `target_interval` RNA interval to expose `editable_positions` permitted DNA positions Rules & boundaries Every candidate pair touching target_interval must become noncanonical. Other folds are not evaluated; no claim of actual accessibility. First/last fixed. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons, Coding DNA to RNA, Canonical RNA base pairs Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.expose_rna_window", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "rna", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." }
Create a guide-escape control`design.guide_escape`	Create a synonymous CDS whose coding RNA no longer closely matches a supplied RNA guide.
Create a guide-escape control The question supplies a guide, permitted DNA-edit locations, and the minimum number of mismatches required between that guide and every potential RNA target window. This can define a matched sequence-recognition control that encodes the same protein. Change only permitted DNA positions using synonymous choices, then compare every same-length RNA window with the guide’s reverse complement. The mismatch minimum must hold everywhere, so removing the original best match must not create another close match nearby. Keep the terminal codons fixed. Practical context Synonymous changes are used to make RNAi-resistant rescue constructs: the same protein is reintroduced from RNA that escapes the knockdown reagent, helping test whether a phenotype came from loss of that protein. The matched-target composition captures the sequence-design step; reagent activity still requires experimental testing. Method reference: RNAi-resistant rescue in a B-cyclin study. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `guide`: `UCACAGUACGAAAAUUGCUUU`; `min_mismatches`: `2`; `editable_positions`: `[6, 9, 12, 15, 18, 21]`. Procedure. Compare the fixed 21-base guide’s reverse complement with every 21-base window of each permitted synonymous candidate. Require at least two mismatches in every window and select the first whole DNA string that passes. Final answer: ATGAAAGCAATATTCGTACTATGA Checking the answer. The candidate must preserve the protein and meet the mismatch minimum at every window. The task makes no claim about actual resistance to an RNA-targeting reagent. Easy example · references supplied Prompt ## Task Rewrite the CDS synonymously so every possible guide-length RNA window meets the required mismatch minimum. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "guide": "UCACAGUACGAAAAUUGCUUU", "min_mismatches": 2, "editable_positions": [6, 9, 12, 15, 18, 21] } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Compare the guide’s reverse complement with every equal-length window of the candidate coding RNA. Each window must differ at at least `min_mismatches` positions. - Check all windows, not only the guide’s original matching location. Use no gaps or wobble pairing. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNA reverse complement RNA complementary bases are A↔U and C↔G. Complement each base, then reverse the sequence to write its complementary partner 5′→3′. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCAATATTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Guide is a designed 21-base reverse complement of minigene RNA bases 4–24, not a validated biological reagent. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously so every possible guide-length RNA window meets the required mismatch minimum. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "guide": "{{guide}}", "min_mismatches": {{min_mismatches}}, "editable_positions": {{editable_positions}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Compare the guide’s reverse complement with every equal-length window of the candidate coding RNA. Each window must differ at at least `min_mismatches` positions. - Check all windows, not only the guide’s original matching location. Use no gaps or wobble pairing. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNA reverse complement RNA complementary bases are A↔U and C↔G. Complement each base, then reverse the sequence to write its complementary partner 5′→3′. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Rewrite the CDS synonymously so every possible guide-length RNA window meets the required mismatch minimum. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "guide": "UCACAGUACGAAAAUUGCUUU", "min_mismatches": 2, "editable_positions": [6, 9, 12, 15, 18, 21] } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Compare the guide’s reverse complement with every equal-length window of the candidate coding RNA. Each window must differ at at least `min_mismatches` positions. - Check all windows, not only the guide’s original matching location. Use no gaps or wobble pairing. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCAATATTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Guide is a designed 21-base reverse complement of minigene RNA bases 4–24, not a validated biological reagent. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Rewrite the CDS synonymously so every possible guide-length RNA window meets the required mismatch minimum. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "guide": "{{guide}}", "min_mismatches": {{min_mismatches}}, "editable_positions": {{editable_positions}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Compare the guide’s reverse complement with every equal-length window of the candidate coding RNA. Each window must differ at at least `min_mismatches` positions. - Check all windows, not only the guide’s original matching location. Use no gaps or wobble pairing. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `guide` 5′→3′ RNA guide `min_mismatches` positive integer `editable_positions` permitted DNA positions Rules & boundaries Every same-length coding-RNA window must have at least min_mismatches to reverse-complement guide. Hamming model only; no efficacy claim. First/last fixed. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons, Coding DNA to RNA, RNA reverse complement Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.guide_escape", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "rna", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." } Used in: Choose an RNA probe and a matched escape control
Build a controlled synonymous panel`design.variant_panel`	Generate several synonymous CDSs with different requested RNA CG counts but the same total number of G/C bases.
Build a controlled synonymous panel The question asks for a matched collection, not one optimized sequence. Every member must encode the original protein. Each has its own target number of adjacent CG occurrences in coding RNA, while all share one required total G+C count. Choose distinct synonymous sequences satisfying those counts exactly. This creates a simple controlled comparison: CG occurrence count varies while protein identity and overall GC composition are held fixed. It does not claim to hold every other RNA property constant. Practical context Matched synonymous panels let experiments compare RNA abundance or protein output while keeping the encoded protein fixed. Here CG count is the intended variable and total GC is additionally controlled. The audit composition verifies those properties; it does not assume other RNA features remain identical. Method reference: Coding-sequence determinants of expression. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `target_cg_counts`: `[0, 1]`; `gc_count`: `6`. Procedure. Enumerate synonymous minigenes with the required total G/C count. Group them by overlapping RNA CG count, then choose a distinct member for each requested count in input order. Apply the whole-panel tie-break. Final answer: ["ATGAAAGCAATATTTGTACTATGA", "ATGAAAGCAATATTCGTATTATGA"] Checking the answer. Every member must meet its assigned CG count and the common GC count, and all must preserve the protein and terminal codons. Easy example · references supplied Prompt ## Task Create a panel of distinct synonymous CDSs with the requested RNA CG counts and a shared DNA GC count. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "target_cg_counts": [0, 1], "gc_count": 6 } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Every returned CDS must be different from every other panel member. - Keep the same order as `target_cg_counts`: the first member must meet the first requested count, the second member the second count, and so on. - Count overlapping CG occurrences in each member’s coding RNA, including occurrences at codon boundaries. - Each complete DNA sequence must contain exactly `gc_count` G/C bases. ## Selection Among complete feasible panels, choose the lexicographically smallest list of DNA strings: compare the first differing member using A<C<G<T. Resolve choices across the whole panel, including distinctness; do not choose each member independently. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ## Answer - Return a list of double-quoted, continuous uppercase synonymous DNA sequences. - Entry i must meet target_cg_counts[i], in that input order. All members must be distinct. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ["ATGAAAGCAATATTTGTACTATGA", "ATGAAAGCAATATTCGTATTATGA"]` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Create a panel of distinct synonymous CDSs with the requested RNA CG counts and a shared DNA GC count. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "target_cg_counts": {{target_cg_counts}}, "gc_count": {{gc_count}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Every returned CDS must be different from every other panel member. - Keep the same order as `target_cg_counts`: the first member must meet the first requested count, the second member the second count, and so on. - Count overlapping CG occurrences in each member’s coding RNA, including occurrences at codon boundaries. - Each complete DNA sequence must contain exactly `gc_count` G/C bases. ## Selection Among complete feasible panels, choose the lexicographically smallest list of DNA strings: compare the first differing member using A<C<G<T. Resolve choices across the whole panel, including distinctness; do not choose each member independently. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ## Answer - Return a list of double-quoted, continuous uppercase synonymous DNA sequences. - Entry i must meet target_cg_counts[i], in that input order. All members must be distinct. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Create a panel of distinct synonymous CDSs with the requested RNA CG counts and a shared DNA GC count. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "target_cg_counts": [0, 1], "gc_count": 6 } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Every returned CDS must be different from every other panel member. - Keep the same order as `target_cg_counts`: the first member must meet the first requested count, the second member the second count, and so on. - Count overlapping CG occurrences in each member’s coding RNA, including occurrences at codon boundaries. - Each complete DNA sequence must contain exactly `gc_count` G/C bases. ## Selection Among complete feasible panels, choose the lexicographically smallest list of DNA strings: compare the first differing member using A<C<G<T. Resolve choices across the whole panel, including distinctness; do not choose each member independently. ## Answer - Return a list of double-quoted, continuous uppercase synonymous DNA sequences. - Entry i must meet target_cg_counts[i], in that input order. All members must be distinct. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ["ATGAAAGCAATATTTGTACTATGA", "ATGAAAGCAATATTCGTATTATGA"]` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Create a panel of distinct synonymous CDSs with the requested RNA CG counts and a shared DNA GC count. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "target_cg_counts": {{target_cg_counts}}, "gc_count": {{gc_count}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Every returned CDS must be different from every other panel member. - Keep the same order as `target_cg_counts`: the first member must meet the first requested count, the second member the second count, and so on. - Count overlapping CG occurrences in each member’s coding RNA, including occurrences at codon boundaries. - Each complete DNA sequence must contain exactly `gc_count` G/C bases. ## Selection Among complete feasible panels, choose the lexicographically smallest list of DNA strings: compare the first differing member using A<C<G<T. Resolve choices across the whole panel, including distinctness; do not choose each member independently. ## Answer - Return a list of double-quoted, continuous uppercase synonymous DNA sequences. - Entry i must meet target_cg_counts[i], in that input order. All members must be distinct. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return a list of double-quoted, continuous uppercase synonymous DNA sequences. Entry i must meet target_cg_counts[i], in that input order. All members must be distinct. Selection Among complete feasible panels, choose the lexicographically smallest list of DNA strings: compare the first differing member using A<C<G<T. Resolve choices across the whole panel, including distinctness; do not choose each member independently. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `target_cg_counts` one desired coding-RNA CG count per panel member `gc_count` same required total G+C count for all members Rules & boundaries Panel member i has target_cg_counts[i] and total gc_count; protein/terminal codons identical; members distinct. Exact target counts, not approximate controls. Among complete feasible panels, choose the lexicographically smallest list of DNA strings: compare the first differing member using A<C<G<T. Resolve choices across the whole panel, including distinctness; do not choose each member independently. Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, NCBI genetic code 11: internal DNA codons, Coding DNA to RNA Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "design.variant_panel", "output_format": "scalar", "scorer": "exact_match", "family": "redesign", "domain": "multi", "objective": "none", "selection_rule": "Among complete feasible panels, choose the lexicographically smallest list of DNA strings: compare the first differing member using A<C<G<T. Resolve choices across the whole panel, including distinctness; do not choose each member independently." } Used in: Build and audit a controlled synonymous panel

07 /

Protein and construct design

Mutations, libraries, tags, and coding modules.

9 tasks

construct task definitions
Task	Description	Expand details
Encode a requested amino-acid substitution`construct.minimal_aa_edit`	Change one requested protein residue by replacing its DNA codon with as few base substitutions as possible.
Encode a requested amino-acid substitution The question identifies an internal residue and the amino acid it should become. Unlike synonymous redesign, this task deliberately changes the protein at that one position. Consider codons encoding the requested amino acid and choose the one differing from the original codon at the fewest bases. Only that codon may change. If several choices use the same number of changes, select the alphabetically first DNA codon. Practical context This is the sequence-design part of site-directed mutagenesis: encode a chosen amino-acid replacement with as few nucleotide substitutions as possible. The protein-editing and cloning composition also removes unwanted assembly sites, turning the residue-level request into a more usable construct sequence. Method reference: Site-directed mutagenesis protocol. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `residue_index`: `3`; `target_aa`: `V`. Procedure. At residue 3, compare each Val codon with the original Ala codon GCA. GTA changes one base and encodes V; choose among minimum-change candidates by the DNA-order rule. Final answer: ATGAAAGTAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA Checking the answer. The intended residue must change, every other base must stay fixed, and the chosen codon must satisfy the minimum-change and tie-breaking rules. Easy example · references supplied Prompt ## Task Change the selected protein residue to `target_aa` with the fewest permitted DNA substitutions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "residue_index": 3, "target_aa": "V" } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Only the selected internal codon may change. Preserve every other DNA base. ## Selection Minimize the total number of changed nucleotides permitted by this task; among equally minimal candidates, return the lexicographically smallest full DNA string (A<C<G<T). ## Reference material ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGTAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Change the selected protein residue to `target_aa` with the fewest permitted DNA substitutions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "residue_index": {{residue_index}}, "target_aa": "{{target_aa}}" } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Only the selected internal codon may change. Preserve every other DNA base. ## Selection Minimize the total number of changed nucleotides permitted by this task; among equally minimal candidates, return the lexicographically smallest full DNA string (A<C<G<T). ## Reference material ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Change the selected protein residue to `target_aa` with the fewest permitted DNA substitutions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "residue_index": 3, "target_aa": "V" } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Only the selected internal codon may change. Preserve every other DNA base. ## Selection Minimize the total number of changed nucleotides permitted by this task; among equally minimal candidates, return the lexicographically smallest full DNA string (A<C<G<T). ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGTAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Change the selected protein residue to `target_aa` with the fewest permitted DNA substitutions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "residue_index": {{residue_index}}, "target_aa": "{{target_aa}}" } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Only the selected internal codon may change. Preserve every other DNA base. ## Selection Minimize the total number of changed nucleotides permitted by this task; among equally minimal candidates, return the lexicographically smallest full DNA string (A<C<G<T). ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Minimize the total number of changed nucleotides permitted by this task; among equally minimal candidates, return the lexicographically smallest full DNA string (A<C<G<T). If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `residue_index` internal residue index `target_aa` one standard amino-acid letter Rules & boundaries Only the selected codon may change; minimize Hamming distance; ties resolved by lexicographic DNA order A<C<G<T. Minimize the total number of changed nucleotides permitted by this task; among equally minimal candidates, return the lexicographically smallest full DNA string (A<C<G<T). Design objective minimize changed nucleotides Scoring `exact_match` · complete parsed answer Reference blocks NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "construct.minimal_aa_edit", "output_format": "scalar", "scorer": "exact_match", "family": "construct", "domain": "protein", "objective": "minimize changed nucleotides", "selection_rule": "Minimize the total number of changed nucleotides permitted by this task; among equally minimal candidates, return the lexicographically smallest full DNA string (A<C<G<T)." } Used in: Encode a protein change and remove cloning sites
Remove a protein motif`construct.remove_peptide_motif`	Modify permitted protein residues to remove a specified peptide pattern, then return DNA encoding the modified product.
Remove a protein motif The question supplies an unwanted literal amino-acid sequence and a list of protein positions that may change, with the allowed replacement residues at each position. It does not ask which natural mutations would preserve function. Choose permitted amino-acid changes until the unwanted pattern is absent from the complete protein. Encode that product in a CDS while retaining its first and terminal stop codons. A change can remove one occurrence but create another, so the final product must be checked globally. Practical context An unwanted protease recognition sequence can interfere with processing an engineered protein. This task removes a specified motif under explicit amino-acid permissions. Motif-to-DNA mapping identifies the encoding codons; the allowed changes must be chosen by the experiment, since motif removal alone cannot establish preserved protein function. Method reference: ExPASy PeptideCutter. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `forbidden_peptide`: `KA`; `allowed_residues`: `{"3": ["A", "G"]}`. Procedure. The minigene encodes MKAIFVL and contains KA at residues 2–3. Use only the permitted A3G substitution to remove that pattern, then choose the first DNA encoding of the allowed final protein. Final answer: ATGAAAGGAATATTCGTACTATGA Checking the answer. The resulting protein must obey every residue permission and contain no forbidden pattern. Retained activity or folding is not part of this task. Easy example · references supplied Prompt ## Task Remove every occurrence of the forbidden peptide using only the allowed amino-acid substitutions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. Numeric keys in position maps are 1-based positions in the complete protein or its aligned codon sequence, including the initiator as position 1. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "forbidden_peptide": "KA", "allowed_residues": { "3": ["A", "G"] } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Remove every literal occurrence of the forbidden peptide. - Only the listed amino-acid positions may change, within their allowed residue sets. All other residues and both terminal codons stay fixed. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGGAATATTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Remove every occurrence of the forbidden peptide using only the allowed amino-acid substitutions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. Numeric keys in position maps are 1-based positions in the complete protein or its aligned codon sequence, including the initiator as position 1. ```json { "cds": "{{cds}}", "forbidden_peptide": "{{forbidden_peptide}}", "allowed_residues": {{allowed_residues}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Remove every literal occurrence of the forbidden peptide. - Only the listed amino-acid positions may change, within their allowed residue sets. All other residues and both terminal codons stay fixed. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Remove every occurrence of the forbidden peptide using only the allowed amino-acid substitutions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. Numeric keys in position maps are 1-based positions in the complete protein or its aligned codon sequence, including the initiator as position 1. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "forbidden_peptide": "KA", "allowed_residues": { "3": ["A", "G"] } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Remove every literal occurrence of the forbidden peptide. - Only the listed amino-acid positions may change, within their allowed residue sets. All other residues and both terminal codons stay fixed. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGGAATATTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Remove every occurrence of the forbidden peptide using only the allowed amino-acid substitutions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. Numeric keys in position maps are 1-based positions in the complete protein or its aligned codon sequence, including the initiator as position 1. ```json { "cds": "{{cds}}", "forbidden_peptide": "{{forbidden_peptide}}", "allowed_residues": {{allowed_residues}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Remove every literal occurrence of the forbidden peptide. - Only the listed amino-acid positions may change, within their allowed residue sets. All other residues and both terminal codons stay fixed. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `forbidden_peptide` literal amino-acid pattern `allowed_residues` allowed residue set per mutable internal position Rules & boundaries Remove all literal peptide hits. All other residues fixed. No claim of retaining fold/function. Choose the declared canonical encoding; terminal codons fixed. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks Annotated CDS initiation, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "construct.remove_peptide_motif", "output_format": "scalar", "scorer": "exact_match", "family": "construct", "domain": "protein", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." }
Build a tagged protein CDS`construct.tag_fusion`	Assemble a CDS encoding a supplied tag and linker at a specified end of the original protein.
Build a tagged protein CDS A tag is a supplied coding module added to a protein construct, often to support purification or detection. The question gives the DNA for the tag and linker, so the model does not invent their sequences. Follow the explicit assembly convention. For an N-terminal tag, retain the original initiating codon, then add tag and linker before the remaining original sense codons. For a C-terminal tag, append linker and tag after the original sense codons. Keep one final stop. Practical context A concrete affinity tag is His6: six histidines that bind nickel affinity material, allowing the attached protein to be captured during purification. The supplied tag and linker DNA encode such modules. The tag-cleavage composition follows a removable tag through processing and calculates the resulting product masses. Method reference: His6-tagged protein purification. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `tag_cds`: `CAT CAT CAT CAT CAT CAT GAA AAT CTG TAT TTT CAG`; `linker_cds`: `GGT GGT TCT`; `placement`: `N`. Procedure. Keep the native initiator, insert the supplied His6–TEV tag and GGS linker, then append the unchanged native sequence from codon 2 through its terminal stop. Final answer: ATGCATCATCATCATCATCATGAAAATCTGTATTTTCAGGGTGGTTCTAAAGCAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA Checking the answer. The constructed DNA must contain the modules in the required order and preserve their reading frame, with no unintended sequence changes. Easy example · references supplied Prompt ## Task Assemble the tagged CDS using the supplied tag, linker, and N- or C-terminal placement. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "tag_cds": "CAT CAT CAT CAT CAT CAT GAA AAT CTG TAT TTT CAG", "linker_cds": "GGT GGT TCT", "placement": "N" } ``` ## Rules - Keep the original first codon and terminal stop codon exactly once in the final construct. - For `placement=N`, insert tag DNA followed by linker DNA immediately after the original first codon. - For `placement=C`, insert linker DNA followed by tag DNA immediately before the original terminal stop. - Apart from the inserted modules, keep every original DNA base unchanged. ## Reference material ### Codon length A codon is a group of three consecutive nucleotides. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGCATCATCATCATCATCATGAAAATCTGTATTTTCAGGGTGGTTCTAAAGCAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The supplied modules encode a His6 tag followed by a TEV site (HHHHHHENLYFQ) and a GGS linker. These are designed additions to native trpL. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Assemble the tagged CDS using the supplied tag, linker, and N- or C-terminal placement. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "tag_cds": "{{tag_cds}}", "linker_cds": "{{linker_cds}}", "placement": "{{placement}}" } ``` ## Rules - Keep the original first codon and terminal stop codon exactly once in the final construct. - For `placement=N`, insert tag DNA followed by linker DNA immediately after the original first codon. - For `placement=C`, insert linker DNA followed by tag DNA immediately before the original terminal stop. - Apart from the inserted modules, keep every original DNA base unchanged. ## Reference material ### Codon length A codon is a group of three consecutive nucleotides. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Assemble the tagged CDS using the supplied tag, linker, and N- or C-terminal placement. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "tag_cds": "CAT CAT CAT CAT CAT CAT GAA AAT CTG TAT TTT CAG", "linker_cds": "GGT GGT TCT", "placement": "N" } ``` ## Rules - Keep the original first codon and terminal stop codon exactly once in the final construct. - For `placement=N`, insert tag DNA followed by linker DNA immediately after the original first codon. - For `placement=C`, insert linker DNA followed by tag DNA immediately before the original terminal stop. - Apart from the inserted modules, keep every original DNA base unchanged. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGCATCATCATCATCATCATGAAAATCTGTATTTTCAGGGTGGTTCTAAAGCAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The supplied modules encode a His6 tag followed by a TEV site (HHHHHHENLYFQ) and a GGS linker. These are designed additions to native trpL. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Assemble the tagged CDS using the supplied tag, linker, and N- or C-terminal placement. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "tag_cds": "{{tag_cds}}", "linker_cds": "{{linker_cds}}", "placement": "{{placement}}" } ``` ## Rules - Keep the original first codon and terminal stop codon exactly once in the final construct. - For `placement=N`, insert tag DNA followed by linker DNA immediately after the original first codon. - For `placement=C`, insert linker DNA followed by tag DNA immediately before the original terminal stop. - Apart from the inserted modules, keep every original DNA base unchanged. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. An empty sequence is the token EMPTY_SEQUENCE. Formal fields and scoring Inputs `cds` valid complete DNA CDS `tag_cds` in-frame DNA codons, no stop; not an independent CDS `linker_cds` in-frame DNA codons, possibly empty `placement` N \| C Rules & boundaries N: original first codon + tag + linker + original sense codons after first + original stop. C: original sense codons + linker + tag + original stop. No other sequence changes. Scoring `exact_match` · complete parsed answer Reference blocks Codon length Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "construct.tag_fusion", "output_format": "scalar", "scorer": "exact_match", "family": "construct", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: Build a tagged construct, cleave it, measure products
Derive TEV-cleaved products`construct.tev_products`	Translate a constructed CDS and determine the protein fragments remaining after the specified TEV cleavage.
Derive TEV-cleaved products TEV protease is a real enzyme derived from tobacco etch virus and widely used to remove engineered affinity tags. ENLYFQG is a common cleavage-site sequence; the cut is ENLYFQ\|G. The benchmark fixes this exact site even though the actual enzyme accepts some additional sequences. Locate all complete recognition sites and split the protein at their cleavage bonds. The G belongs to the downstream fragment. Return every fragment in its original order rather than automatically discarding the tag or reporting only a preferred product. Practical context TEV protease is used to remove affinity tags after protein purification. Predicting the products establishes what tag sequence or cleavage scar remains on the target. The cleavage-and-mass composition uses those exact sequences when deriving an expected identity readout. Method reference: Removal of affinity tags with TEV protease. Task Outline Inputs. `cds`: `ATG CAT CAT CAT CAT CAT CAT GAA AAT CTG TAT TTT CAG GGT GGT TCT AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`. Procedure. Translate the tagged native construct. Find ENLYFQG and cut ENLYFQ\|G, leaving the GGS linker on the downstream fragment. Retain the two nonempty protein fragments in their original order. Final answer: ["MHHHHHHENLYFQ", "GGSKAIFVLKGWWRTS"] Checking the answer. The fragments must reconstruct the original protein when joined and place each cleavage at the stated bond. Easy example · references supplied Prompt ## Task List the protein fragments produced by complete cleavage under the benchmark TEV rule. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG CAT CAT CAT CAT CAT CAT GAA AAT CTG TAT TTT CAG GGT GGT TCT AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - For this benchmark, recognize only the exact sequence `ENLYFQG` and cut it as `ENLYFQ\|G`. Keep G on the downstream fragment. - Cut every occurrence and return the nonempty fragments in their original protein order. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return a list of double-quoted, continuous uppercase peptide sequences, in N-to-C fragment order. Keep every nonempty fragment, including repeated sequences. - With no cleavage, return a one-item list containing the complete protein. An empty protein gives []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ["MHHHHHHENLYFQ", "GGSKAIFVLKGWWRTS"]` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL after inserting His6–TEV-site tag DNA CATCATCATCATCATCATGAAAATCTGTATTTTCAG and GGS linker GGTGGTTCT immediately after its initiator. This is the same construct produced by tag_fusion and tag_cleave_mass. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task List the protein fragments produced by complete cleavage under the benchmark TEV rule. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - For this benchmark, recognize only the exact sequence `ENLYFQG` and cut it as `ENLYFQ\|G`. Keep G on the downstream fragment. - Cut every occurrence and return the nonempty fragments in their original protein order. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return a list of double-quoted, continuous uppercase peptide sequences, in N-to-C fragment order. Keep every nonempty fragment, including repeated sequences. - With no cleavage, return a one-item list containing the complete protein. An empty protein gives []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task List the protein fragments produced by complete cleavage under the benchmark TEV rule. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG CAT CAT CAT CAT CAT CAT GAA AAT CTG TAT TTT CAG GGT GGT TCT AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA" } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - For this benchmark, recognize only the exact sequence `ENLYFQG` and cut it as `ENLYFQ\|G`. Keep G on the downstream fragment. - Cut every occurrence and return the nonempty fragments in their original protein order. ## Answer - Return a list of double-quoted, continuous uppercase peptide sequences, in N-to-C fragment order. Keep every nonempty fragment, including repeated sequences. - With no cleavage, return a one-item list containing the complete protein. An empty protein gives []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ["MHHHHHHENLYFQ", "GGSKAIFVLKGWWRTS"]` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL after inserting His6–TEV-site tag DNA CATCATCATCATCATCATGAAAATCTGTATTTTCAG and GGS linker GGTGGTTCT immediately after its initiator. This is the same construct produced by tag_fusion and tag_cleave_mass. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task List the protein fragments produced by complete cleavage under the benchmark TEV rule. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}" } ``` ## Rules - Translate the CDS using NCBI genetic code 11, stopping at the first in-frame stop. - For this benchmark, recognize only the exact sequence `ENLYFQG` and cut it as `ENLYFQ\|G`. Keep G on the downstream fragment. - Cut every occurrence and return the nonempty fragments in their original protein order. ## Answer - Return a list of double-quoted, continuous uppercase peptide sequences, in N-to-C fragment order. Keep every nonempty fragment, including repeated sequences. - With no cleavage, return a one-item list containing the complete protein. An empty protein gives []. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return a list of double-quoted, continuous uppercase peptide sequences, in N-to-C fragment order. Keep every nonempty fragment, including repeated sequences. With no cleavage, return a one-item list containing the complete protein. An empty protein gives []. Formal fields and scoring Inputs `protein` protein derived from a constructed CDS Rules & boundaries Rule tev_ENLYFQ_G_v1; cleave every complete ENLYFQG match after Q; include G downstream. Scoring `exact_match` · complete parsed answer Reference blocks Annotated CDS initiation, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "construct.tev_products", "output_format": "scalar", "scorer": "exact_match", "family": "construct", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: Build a tagged construct, cleave it, measure products
Enumerate a degenerate codon library`library.codon_products`	Expand an ambiguous internal DNA codon and count how many concrete codons encode each amino acid or stop.
Enumerate a degenerate codon library Degenerate oligonucleotide synthesis mixes nucleotide choices at selected positions, producing a pool of DNA sequences for mutagenesis experiments. IUPAC notation describes those choices compactly: N, for example, permits A/C/G/T. The question asks which protein alternatives such a mixed codon covers at one CDS position. Expand all possible base combinations and translate each codon internally. Count concrete codons per amino acid, rather than reporting each amino acid once. These are combinatorial counts, not measured library frequencies or expression levels. Practical context Before ordering a degenerate oligonucleotide, library designers need to know which amino acids it covers, how redundantly it encodes them, and whether it includes stops. Library design and verification uses this enumeration to audit a proposed mixed-codon design. Method reference: SwiftLib. Task Outline Inputs. `cds`: `ATG AAA NNK ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `codon_index`: `3`. Procedure. Expand NNK into 4×4×2 = 32 distinct DNA codons and translate each internally. Increment the corresponding amino-acid or STOP count once per concrete codon, including outcomes outside the native Ala residue. Final answer: { "A": 2, "C": 1, "D": 1, "E": 1, "F": 1, "G": 2, "H": 1, "I": 1, "K": 1, "L": 3, "M": 1, "N": 1, "P": 2, "Q": 1, "R": 3, "S": 3, "STOP": 1, "T": 2, "V": 2, "W": 1, "Y": 1 } Checking the answer. The residue and stop counts must account for every concrete codon in the expansion exactly once. Easy example · references supplied Prompt ## Task Count how many concrete DNA codons in the selected IUPAC codon’s expansion encode each amino acid or stop. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA NNK ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "codon_index": 3 } ``` ## Rules - Inspect only the internal codon at `codon_index`. Expand its three IUPAC positions independently. - Translate each distinct concrete codon with its ordinary internal NCBI genetic code 11 meaning. - For each resulting amino acid or stop, count the concrete codons that produce it. These are codon counts, not probabilities. ## Reference material ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Map each produced uppercase one-letter amino-acid symbol, or "STOP", to the number of distinct concrete codons encoding that outcome in the selected IUPAC codon’s expansion. - Include every positive count, including unintended amino acids and stops. The counts sum to the expansion size; they are codon multiplicities rather than probabilities or full-protein residue counts. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<one-letter amino-acid symbol or STOP>": <positive integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "A": 2, "C": 1, "D": 1, "E": 1, "F": 1, "G": 2, "H": 1, "I": 1, "K": 1, "L": 3, "M": 1, "N": 1, "P": 2, "Q": 1, "R": 3, "S": 3, "STOP": 1, "T": 2, "V": 2, "W": 1, "Y": 1 }` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Replace native Ala codon 3 with NNK to represent a designed 32-codon saturation library. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Count how many concrete DNA codons in the selected IUPAC codon’s expansion encode each amino acid or stop. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "codon_index": {{codon_index}} } ``` ## Rules - Inspect only the internal codon at `codon_index`. Expand its three IUPAC positions independently. - Translate each distinct concrete codon with its ordinary internal NCBI genetic code 11 meaning. - For each resulting amino acid or stop, count the concrete codons that produce it. These are codon counts, not probabilities. ## Reference material ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Map each produced uppercase one-letter amino-acid symbol, or "STOP", to the number of distinct concrete codons encoding that outcome in the selected IUPAC codon’s expansion. - Include every positive count, including unintended amino acids and stops. The counts sum to the expansion size; they are codon multiplicities rather than probabilities or full-protein residue counts. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<one-letter amino-acid symbol or STOP>": <positive integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Count how many concrete DNA codons in the selected IUPAC codon’s expansion encode each amino acid or stop. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA NNK ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "codon_index": 3 } ``` ## Rules - Inspect only the internal codon at `codon_index`. Expand its three IUPAC positions independently. - Translate each distinct concrete codon with its ordinary internal NCBI genetic code 11 meaning. - For each resulting amino acid or stop, count the concrete codons that produce it. These are codon counts, not probabilities. ## Answer - Map each produced uppercase one-letter amino-acid symbol, or "STOP", to the number of distinct concrete codons encoding that outcome in the selected IUPAC codon’s expansion. - Include every positive count, including unintended amino acids and stops. The counts sum to the expansion size; they are codon multiplicities rather than probabilities or full-protein residue counts. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<one-letter amino-acid symbol or STOP>": <positive integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "A": 2, "C": 1, "D": 1, "E": 1, "F": 1, "G": 2, "H": 1, "I": 1, "K": 1, "L": 3, "M": 1, "N": 1, "P": 2, "Q": 1, "R": 3, "S": 3, "STOP": 1, "T": 2, "V": 2, "W": 1, "Y": 1 }` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Replace native Ala codon 3 with NNK to represent a designed 32-codon saturation library. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Count how many concrete DNA codons in the selected IUPAC codon’s expansion encode each amino acid or stop. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "codon_index": {{codon_index}} } ``` ## Rules - Inspect only the internal codon at `codon_index`. Expand its three IUPAC positions independently. - Translate each distinct concrete codon with its ordinary internal NCBI genetic code 11 meaning. - For each resulting amino acid or stop, count the concrete codons that produce it. These are codon counts, not probabilities. ## Answer - Map each produced uppercase one-letter amino-acid symbol, or "STOP", to the number of distinct concrete codons encoding that outcome in the selected IUPAC codon’s expansion. - Include every positive count, including unintended amino acids and stops. The counts sum to the expansion size; they are codon multiplicities rather than probabilities or full-protein residue counts. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "<one-letter amino-acid symbol or STOP>": <positive integer count> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` Map each produced uppercase one-letter amino-acid symbol, or "STOP", to the number of distinct concrete codons encoding that outcome in the selected IUPAC codon’s expansion. Include every positive count, including unintended amino acids and stops. The counts sum to the expansion size; they are codon multiplicities rather than probabilities or full-protein residue counts. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "<one-letter amino-acid symbol or STOP>": <positive integer count> } Formal fields and scoring Inputs `cds` complete CDS with one internal IUPAC codon `codon_index` selected internal codon Rules & boundaries Expand independently and count codon multiplicities; not biological abundance or synthesis yield. Scoring `exact_match` · complete parsed answer Reference blocks IUPAC DNA base sets, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "library.codon_products", "output_format": "json", "scorer": "exact_match", "family": "construct", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: Design a codon library and verify its coverage
Design a focused degenerate codon`library.design_codon`	Choose an ambiguous DNA codon that covers requested amino acids without encoding stops, while limiting unwanted alternatives.
Design a focused degenerate codon The question supplies the amino acids desired at one internal CDS position. One IUPAC triplet may cover those choices compactly, but its expansion can also include additional amino acids. Search triplets whose concrete codons cover every requested residue and none of the stops. Prefer the fewest extra amino-acid types, then the fewest concrete codons. A final alphabetical tie-break makes the answer unique. If no triplet can satisfy stop-free coverage, return IMPOSSIBLE. Practical context Focused mutagenesis libraries target selected amino acids while limiting how many variants must be screened. Degenerate-codon optimization tools such as SwiftLib address that trade-off. The design-and-audit composition checks the resulting codon mixture’s actual coverage instead of relying on its intended target set. Method reference: SwiftLib. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `codon_index`: `3`; `target_residues`: `["A", "G", "V"]`. Procedure. Enumerate all 15³ IUPAC triplets, expand each into concrete codons, and discard stops or incomplete A/G/V coverage. Minimize extra amino-acid types, then expansion size, then the IUPAC string. Final answer: GBA Checking the answer. Coverage and stop avoidance are mandatory. The answer must also achieve the declared ordering of the optimization criteria. Easy example · references supplied Prompt ## Task Design an IUPAC codon that covers every target amino acid and cannot encode a stop. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "codon_index": 3, "target_residues": ["A", "G", "V"] } ``` ## Rules - Choose one three-symbol IUPAC DNA codon for the selected internal position. - Its concrete expansions must encode every residue in `target_residues` and must never encode a stop under NCBI genetic code 11. - Use the ranking below to choose among qualifying IUPAC codons. ## Selection Require target-residue coverage and no stops. Minimize extra amino-acid types, then the number of concrete codons, then the IUPAC triplet in ASCII order A<B<C<D<G<H<K<M<N<R<S<T<V<W<Y. Compare the symbols themselves, not their expanded nucleotide sets. ## Reference material ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return the selected uppercase three-character IUPAC DNA triplet. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: GBA` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Design an IUPAC codon that covers every target amino acid and cannot encode a stop. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "codon_index": {{codon_index}}, "target_residues": {{target_residues}} } ``` ## Rules - Choose one three-symbol IUPAC DNA codon for the selected internal position. - Its concrete expansions must encode every residue in `target_residues` and must never encode a stop under NCBI genetic code 11. - Use the ranking below to choose among qualifying IUPAC codons. ## Selection Require target-residue coverage and no stops. Minimize extra amino-acid types, then the number of concrete codons, then the IUPAC triplet in ASCII order A<B<C<D<G<H<K<M<N<R<S<T<V<W<Y. Compare the symbols themselves, not their expanded nucleotide sets. ## Reference material ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return the selected uppercase three-character IUPAC DNA triplet. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Design an IUPAC codon that covers every target amino acid and cannot encode a stop. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "codon_index": 3, "target_residues": ["A", "G", "V"] } ``` ## Rules - Choose one three-symbol IUPAC DNA codon for the selected internal position. - Its concrete expansions must encode every residue in `target_residues` and must never encode a stop under NCBI genetic code 11. - Use the ranking below to choose among qualifying IUPAC codons. ## Selection Require target-residue coverage and no stops. Minimize extra amino-acid types, then the number of concrete codons, then the IUPAC triplet in ASCII order A<B<C<D<G<H<K<M<N<R<S<T<V<W<Y. Compare the symbols themselves, not their expanded nucleotide sets. ## Answer - Return the selected uppercase three-character IUPAC DNA triplet. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: GBA` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Design an IUPAC codon that covers every target amino acid and cannot encode a stop. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "codon_index": {{codon_index}}, "target_residues": {{target_residues}} } ``` ## Rules - Choose one three-symbol IUPAC DNA codon for the selected internal position. - Its concrete expansions must encode every residue in `target_residues` and must never encode a stop under NCBI genetic code 11. - Use the ranking below to choose among qualifying IUPAC codons. ## Selection Require target-residue coverage and no stops. Minimize extra amino-acid types, then the number of concrete codons, then the IUPAC triplet in ASCII order A<B<C<D<G<H<K<M<N<R<S<T<V<W<Y. Compare the symbols themselves, not their expanded nucleotide sets. ## Answer - Return the selected uppercase three-character IUPAC DNA triplet. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return the selected uppercase three-character IUPAC DNA triplet. Selection Require target-residue coverage and no stops. Minimize extra amino-acid types, then the number of concrete codons, then the IUPAC triplet in ASCII order A<B<C<D<G<H<K<M<N<R<S<T<V<W<Y. Compare the symbols themselves, not their expanded nucleotide sets. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `codon_index` internal position `target_residues` nonempty set of standard amino-acid symbols Rules & boundaries Its expansion must cover every target residue and contain no STOP. Minimize extra residue types, then number of concrete codons, then ASCII lexicographic triplet. Search all 15^3 IUPAC triplets. Require target-residue coverage and no stops. Minimize extra amino-acid types, then the number of concrete codons, then the IUPAC triplet in ASCII order A<B<C<D<G<H<K<M<N<R<S<T<V<W<Y. Compare the symbols themselves, not their expanded nucleotide sets. Design objective lexicographic objective: extra residue types, concrete codons, triplet Scoring `exact_match` · complete parsed answer Reference blocks IUPAC DNA base sets, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "library.design_codon", "output_format": "scalar", "scorer": "exact_match", "family": "construct", "domain": "protein", "objective": "lexicographic objective: extra residue types, concrete codons, triplet", "selection_rule": "Require target-residue coverage and no stops. Minimize extra amino-acid types, then the number of concrete codons, then the IUPAC triplet in ASCII order A<B<C<D<G<H<K<M<N<R<S<T<V<W<Y. Compare the symbols themselves, not their expanded nucleotide sets." } Used in: Design a codon library and verify its coverage
Delete a protein interval`construct.inframe_deletion`	Remove a specified range of protein residues by deleting exactly the corresponding complete DNA codons.
Delete a protein interval The question identifies an internal residue interval in the protein encoded by a CDS. The goal is to remove that segment without shifting the grouping of downstream codons. Map the residue interval to DNA, delete those complete codons, and leave all other DNA unchanged. The interval excludes the initiating residue, and the original terminal stop is retained. Practical context Deletion constructs can test the contribution of an explicitly chosen protein region. This task turns those residue boundaries into an in-frame DNA deletion. It prepares the sequence for such a comparison; deciding whether the shortened protein still folds or functions requires evidence beyond the deletion operation. Method reference: Site-directed mutagenesis protocol. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `aa_interval`: `[8, 9]`. Procedure. Remove complete codons 8–9, DNA bases 22–27, without changing the order or identities of any remaining bases. Keep the original initiator and terminal TGA. Final answer: ATGAAAGCAATTTTCGTACTGTGGTGGCGCACTTCCTGA Checking the answer. Exactly the requested codons must be deleted; synonymous changes elsewhere are not allowed in this task. Easy example Prompt ## Task Delete the complete codons encoding `aa_interval` and return the remaining CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "aa_interval": [8, 9] } ``` ## Rules - The requested residue interval is inclusive and lies within the internal sense region. Remove all three DNA bases of each selected codon. - Keep every remaining base, including the original first codon and terminal stop, unchanged and in its original order. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCAATTTTCGTACTGTGGTGGCGCACTTCCTGA` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Delete the complete codons encoding `aa_interval` and return the remaining CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "aa_interval": {{aa_interval}} } ``` ## Rules - The requested residue interval is inclusive and lies within the internal sense region. Remove all three DNA bases of each selected codon. - Keep every remaining base, including the original first codon and terminal stop, unchanged and in its original order. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example Prompt ## Task Delete the complete codons encoding `aa_interval` and return the remaining CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "aa_interval": [8, 9] } ``` ## Rules - The requested residue interval is inclusive and lies within the internal sense region. Remove all three DNA bases of each selected codon. - Keep every remaining base, including the original first codon and terminal stop, unchanged and in its original order. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCAATTTTCGTACTGTGGTGGCGCACTTCCTGA` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Delete the complete codons encoding `aa_interval` and return the remaining CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "aa_interval": {{aa_interval}} } ``` ## Rules - The requested residue interval is inclusive and lies within the internal sense region. Remove all three DNA bases of each selected codon. - Keep every remaining base, including the original first codon and terminal stop, unchanged and in its original order. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. An empty sequence is the token EMPTY_SEQUENCE. Formal fields and scoring Inputs `cds` valid complete DNA CDS `aa_interval` internal residue interval; excludes initiator Rules & boundaries Delete exactly the encoding complete codons, preserving remaining DNA and terminal stop. Scoring `exact_match` · complete parsed answer Reference blocks None required. Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "construct.inframe_deletion", "output_format": "scalar", "scorer": "exact_match", "family": "construct", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." }
Join coding modules`construct.join_modules`	Join two supplied CDSs with a coding linker to make one continuous fusion-protein CDS.
Join coding modules The question gives two complete coding sequences and a linker made of complete, stop-free codons. Joining the DNA naively would retain the first CDS’s stop and prevent translation into the second module. Remove the first CDS’s terminal stop and the second CDS’s initiating codon. Join the remaining first module, linker, and second module, retaining the first initiator and the second terminal stop. This fixes precisely which residues appear at the junction. Practical context Fusion constructs combine a target protein with reporters, tags, or other supplied coding modules. Correct reading frame and stop handling are essential at each join. The junction-audit composition then checks whether assembly created an unwanted peptide motif across a module boundary. Method reference: Addgene bacterial expression collection. Task Outline Inputs. `cds_a`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `cds_b`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `linker_cds`: `GGT GGT TCT`. Procedure. Remove only the full-length module’s final TGA and the minigene module’s initial ATG. Join the remaining modules around the supplied GGS linker, retaining the first module’s initiator and the second’s stop. Final answer: ATGAAAGCAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCGGTGGTTCTAAAGCAATTTTCGTACTGTGA Checking the answer. The fused DNA must follow that exact junction convention and preserve all remaining module bases. Easy example · references supplied Prompt ## Task Join CDS A, the linker, and CDS B into a single CDS under the assembly rule below. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_a": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "cds_b": "ATG AAA GCA ATT TTC GTA CTG TGA", "linker_cds": "GGT GGT TCT" } ``` ## Rules - Remove only the terminal stop codon from A and the first codon from B. - Concatenate the remaining A sequence, `linker_cds`, and the remaining B sequence in that order. - Keep A’s first codon and B’s terminal stop. Do not remove or alter any other bases. ## Reference material ### Codon length A codon is a group of three consecutive nucleotides. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCGGTGGTTCTAAAGCAATTTTCGTACTGTGA` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Join native full-length trpL to the explicitly truncated trpL minigene using a designed GGS linker. This is a constructed tandem fusion, not a natural gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Join CDS A, the linker, and CDS B into a single CDS under the assembly rule below. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_a": "{{cds_a}}", "cds_b": "{{cds_b}}", "linker_cds": "{{linker_cds}}" } ``` ## Rules - Remove only the terminal stop codon from A and the first codon from B. - Concatenate the remaining A sequence, `linker_cds`, and the remaining B sequence in that order. - Keep A’s first codon and B’s terminal stop. Do not remove or alter any other bases. ## Reference material ### Codon length A codon is a group of three consecutive nucleotides. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Join CDS A, the linker, and CDS B into a single CDS under the assembly rule below. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_a": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "cds_b": "ATG AAA GCA ATT TTC GTA CTG TGA", "linker_cds": "GGT GGT TCT" } ``` ## Rules - Remove only the terminal stop codon from A and the first codon from B. - Concatenate the remaining A sequence, `linker_cds`, and the remaining B sequence in that order. - Keep A’s first codon and B’s terminal stop. Do not remove or alter any other bases. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCGGTGGTTCTAAAGCAATTTTCGTACTGTGA` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Join native full-length trpL to the explicitly truncated trpL minigene using a designed GGS linker. This is a constructed tandem fusion, not a natural gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Join CDS A, the linker, and CDS B into a single CDS under the assembly rule below. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_a": "{{cds_a}}", "cds_b": "{{cds_b}}", "linker_cds": "{{linker_cds}}" } ``` ## Rules - Remove only the terminal stop codon from A and the first codon from B. - Concatenate the remaining A sequence, `linker_cds`, and the remaining B sequence in that order. - Keep A’s first codon and B’s terminal stop. Do not remove or alter any other bases. ## Answer - Return one continuous uppercase DNA sequence. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. An empty sequence is the token EMPTY_SEQUENCE. Formal fields and scoring Inputs `cds_a` complete DNA CDS `cds_b` complete DNA CDS `linker_cds` in-frame stop-free DNA codons, possibly empty Rules & boundaries Drop A terminal stop and B initial codon. Join A sense DNA + linker + B after initiator including B stop. Retain A initiator. Exact convention, no automatic cleavage. Scoring `exact_match` · complete parsed answer Reference blocks Codon length Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "construct.join_modules", "output_format": "scalar", "scorer": "exact_match", "family": "construct", "domain": "protein", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: Assemble coding modules and locate junction motifs
Design an RNA-editing-dependent product`construct.edit_aware_encoding`	Design a synonymous CDS whose coding RNA produces a requested different protein after explicitly specified RNA edits.
Design an RNA-editing-dependent product The question supplies an original CDS, fixed RNA editing events, a desired post-edit protein, and the DNA positions that may change. Before editing, the designed DNA must still encode the original protein. Choose permitted synonymous codons that make those particular RNA edits yield the desired product. Check two branches: translation before editing and translation after transcription plus editing. Editing sites and enzyme behavior are given, not predicted. Exercise scope ADAR enzymes convert RNA adenosine to inosine, which is often read as G during translation. The example represents that recoding as A→G and asks for a synonymous DNA encoding with different unedited and edited proteins. This is a two-state sequence exercise: the editing event is assumed, not evidence that ADAR will act at the chosen site. Method reference: Programmable editing by endogenous ADAR. Task Outline Inputs. `cds`: `ATG AAA TCT ATT TTC GTA CTG TGA`; `rna_edits`: `[{"replace": [7, 7, "G"]}]`; `target_protein`: `MKGIFVL`; `editable_positions`: `[7, 8]`. Procedure. The unedited codon 3 must still encode Ser. Change TCT to AGT using the two permitted DNA positions; the supplied RNA edit then changes AGU to GGU, encoding the required Gly. Final answer: ATGAAAAGTATTTTCGTACTGTGA Checking the answer. One designed CDS must pass both protein requirements and every DNA-edit restriction. An answer that changes the unedited protein is invalid. Easy example · references supplied Prompt ## Task Design a synonymous CDS that encodes the original protein before RNA editing and `target_protein` after the specified RNA edits. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA TCT ATT TTC GTA CTG TGA", "rna_edits": [ { "replace": [7, 7, "G"] } ], "target_protein": "MKGIFVL", "editable_positions": [7, 8] } ``` ## Rules - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Transcribe the candidate CDS, apply exactly the supplied RNA edits in original RNA coordinates, and translate the edited RNA with NCBI genetic code 11. - The unedited candidate must encode the original complete protein, and the edited RNA must encode exactly `target_protein`. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Initiation of coding RNA At the first RNA codon, UUG, CUG, AUU, AUC, AUA, AUG, and GUG encode initiating M under NCBI genetic code 11. If that first codon is not an allowed initiator, use its ordinary internal assignment. ### NCBI genetic code 11: internal RNA codons Each internal RNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text UUU=F UUC=F UUA=L UUG=L UCU=S UCC=S UCA=S UCG=S UAU=Y UAC=Y UAA=STOP UAG=STOP UGU=C UGC=C UGA=STOP UGG=W CUU=L CUC=L CUA=L CUG=L CCU=P CCC=P CCA=P CCG=P CAU=H CAC=H CAA=Q CAG=Q CGU=R CGC=R CGA=R CGG=R AUU=I AUC=I AUA=I AUG=M ACU=T ACC=T ACA=T ACG=T AAU=N AAC=N AAA=K AAG=K AGU=S AGC=S AGA=R AGG=R GUU=V GUC=V GUA=V GUG=V GCU=A GCC=A GCA=A GCG=A GAU=D GAC=D GAA=E GAG=E GGU=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAAGTATTTTCGTACTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The minigene has A3S (GCA→TCT). The supplied RNA edit sets base 7 to G; choose a synonymous Ser encoding so that this edit yields Gly. No naturally observed editing is claimed. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Design a synonymous CDS that encodes the original protein before RNA editing and `target_protein` after the specified RNA edits. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "rna_edits": {{rna_edits}}, "target_protein": "{{target_protein}}", "editable_positions": {{editable_positions}} } ``` ## Rules - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Transcribe the candidate CDS, apply exactly the supplied RNA edits in original RNA coordinates, and translate the edited RNA with NCBI genetic code 11. - The unedited candidate must encode the original complete protein, and the edited RNA must encode exactly `target_protein`. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Initiation of coding RNA At the first RNA codon, UUG, CUG, AUU, AUC, AUA, AUG, and GUG encode initiating M under NCBI genetic code 11. If that first codon is not an allowed initiator, use its ordinary internal assignment. ### NCBI genetic code 11: internal RNA codons Each internal RNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text UUU=F UUC=F UUA=L UUG=L UCU=S UCC=S UCA=S UCG=S UAU=Y UAC=Y UAA=STOP UAG=STOP UGU=C UGC=C UGA=STOP UGG=W CUU=L CUC=L CUA=L CUG=L CCU=P CCC=P CCA=P CCG=P CAU=H CAC=H CAA=Q CAG=Q CGU=R CGC=R CGA=R CGG=R AUU=I AUC=I AUA=I AUG=M ACU=T ACC=T ACA=T ACG=T AAU=N AAC=N AAA=K AAG=K AGU=S AGC=S AGA=R AGG=R GUU=V GUC=V GUA=V GUG=V GCU=A GCC=A GCA=A GCG=A GAU=D GAC=D GAA=E GAG=E GGU=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Design a synonymous CDS that encodes the original protein before RNA editing and `target_protein` after the specified RNA edits. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA TCT ATT TTC GTA CTG TGA", "rna_edits": [ { "replace": [7, 7, "G"] } ], "target_protein": "MKGIFVL", "editable_positions": [7, 8] } ``` ## Rules - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Transcribe the candidate CDS, apply exactly the supplied RNA edits in original RNA coordinates, and translate the edited RNA with NCBI genetic code 11. - The unedited candidate must encode the original complete protein, and the edited RNA must encode exactly `target_protein`. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAAGTATTTTCGTACTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The minigene has A3S (GCA→TCT). The supplied RNA edit sets base 7 to G; choose a synonymous Ser encoding so that this edit yields Gly. No naturally observed editing is claimed. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Design a synonymous CDS that encodes the original protein before RNA editing and `target_protein` after the specified RNA edits. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "rna_edits": {{rna_edits}}, "target_protein": "{{target_protein}}", "editable_positions": {{editable_positions}} } ``` ## Rules - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Transcribe the candidate CDS, apply exactly the supplied RNA edits in original RNA coordinates, and translate the edited RNA with NCBI genetic code 11. - The unedited candidate must encode the original complete protein, and the edited RNA must encode exactly `target_protein`. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` valid complete DNA CDS `rna_edits` explicit A/C/G/U substitutions at fixed positions `target_protein` desired post-edit amino-acid string `editable_positions` permitted DNA changes Rules & boundaries Pre-edit protein must equal original; post-transcription specified edits must yield target_protein. No edit-site discovery or enzyme efficiency model. First/last codons fixed. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks Synonymous codons, Coding DNA to RNA, Initiation of coding RNA, NCBI genetic code 11: internal RNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "construct.edit_aware_encoding", "output_format": "scalar", "scorer": "exact_match", "family": "construct", "domain": "rna", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." } Used in: Design and verify an RNA-editing-dependent product

08 /

Assay design

CDS-targeted primers and discriminating probe panels.

3 tasks

assay task definitions
Task	Description	Expand details
Choose CDS-internal PCR primers`assay.primer_pair`	Choose forward and reverse DNA primers for a specified interval entirely within a supplied CDS.
Choose CDS-internal PCR primers The question fixes the desired amplified interval and primer length. It also supplies acceptable GC fractions. The primers must bind the intended ends in opposing directions; no genomic flanks need to be supplied. Take the first primer-length bases of the interval for the forward primer. Reverse-complement the last primer-length bases for the reverse primer. Check both composition limits and uniqueness of their intended oriented binding sites within this CDS. Practical context Correct primer orientation and target endpoints are prerequisites for amplifying a CDS segment. This task isolates that sequence-selection step. Practical tools such as Primer3 also evaluate melting temperatures, self-complementarity, and interactions between primers when choosing a pair for an assay. Method reference: Primer3. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `amplicon_interval`: `[1, 45]`; `primer_length`: `18`; `primer_gc_bounds`: `[0.25, 0.65]`. Procedure. Take native bases 1–18 as the forward primer. Reverse-complement bases 28–45 for the reverse primer. Check both GC fractions and confirm that each binding sequence occurs once in the supplied CDS. Final answer: { "status": "ok", "forward": "ATGAAAGCAATTTTCGTA", "reverse": "TCAGGAAGTGCGCCACCA" } Checking the answer. Both sequences must have the correct orientation and interval endpoints and satisfy the supplied constraints. Melting behavior and genome-wide specificity are outside this version. Easy example · references supplied Prompt ## Task Return the fixed-length forward and reverse primers for the supplied CDS amplicon. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "amplicon_interval": [1, 45], "primer_length": 18, "primer_gc_bounds": [0.25, 0.65] } ``` ## Rules - Use the exact endpoints in `amplicon_interval`; do not move them to find better primers. - The forward primer is the first `primer_length` DNA bases of the amplicon. The reverse primer is the reverse complement of its last `primer_length` bases. Write both primers 5′→3′. - Both primers must meet the inclusive `primer_gc_bounds`. - The forward primer sequence must occur once on the supplied coding strand; the reverse primer’s reverse complement must occur once. Check uniqueness only within this CDS. - If either fixed primer fails a check, return the specified failure object rather than choosing different endpoints. ## Reference material ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ### GC fraction GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured. ## Answer - `forward` is the amplicon’s first primer_length DNA bases. `reverse` is the reverse complement of its final primer_length bases. Write both as continuous uppercase 5′→3′ DNA strings. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "forward": <DNA string>, "reverse": <DNA string> } ``` If either fixed primer fails the supplied checks, return: ```json { "status": "impossible", "forward": null, "reverse": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "status": "ok", "forward": "ATGAAAGCAATTTTCGTA", "reverse": "TCAGGAAGTGCGCCACCA" }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Return the fixed-length forward and reverse primers for the supplied CDS amplicon. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "amplicon_interval": {{amplicon_interval}}, "primer_length": {{primer_length}}, "primer_gc_bounds": {{primer_gc_bounds}} } ``` ## Rules - Use the exact endpoints in `amplicon_interval`; do not move them to find better primers. - The forward primer is the first `primer_length` DNA bases of the amplicon. The reverse primer is the reverse complement of its last `primer_length` bases. Write both primers 5′→3′. - Both primers must meet the inclusive `primer_gc_bounds`. - The forward primer sequence must occur once on the supplied coding strand; the reverse primer’s reverse complement must occur once. Check uniqueness only within this CDS. - If either fixed primer fails a check, return the specified failure object rather than choosing different endpoints. ## Reference material ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ### GC fraction GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured. ## Answer - `forward` is the amplicon’s first primer_length DNA bases. `reverse` is the reverse complement of its final primer_length bases. Write both as continuous uppercase 5′→3′ DNA strings. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "forward": <DNA string>, "reverse": <DNA string> } ``` If either fixed primer fails the supplied checks, return: ```json { "status": "impossible", "forward": null, "reverse": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Return the fixed-length forward and reverse primers for the supplied CDS amplicon. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "amplicon_interval": [1, 45], "primer_length": 18, "primer_gc_bounds": [0.25, 0.65] } ``` ## Rules - Use the exact endpoints in `amplicon_interval`; do not move them to find better primers. - The forward primer is the first `primer_length` DNA bases of the amplicon. The reverse primer is the reverse complement of its last `primer_length` bases. Write both primers 5′→3′. - Both primers must meet the inclusive `primer_gc_bounds`. - The forward primer sequence must occur once on the supplied coding strand; the reverse primer’s reverse complement must occur once. Check uniqueness only within this CDS. - If either fixed primer fails a check, return the specified failure object rather than choosing different endpoints. ## Answer - `forward` is the amplicon’s first primer_length DNA bases. `reverse` is the reverse complement of its final primer_length bases. Write both as continuous uppercase 5′→3′ DNA strings. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "forward": <DNA string>, "reverse": <DNA string> } ``` If either fixed primer fails the supplied checks, return: ```json { "status": "impossible", "forward": null, "reverse": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "status": "ok", "forward": "ATGAAAGCAATTTTCGTA", "reverse": "TCAGGAAGTGCGCCACCA" }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Return the fixed-length forward and reverse primers for the supplied CDS amplicon. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "amplicon_interval": {{amplicon_interval}}, "primer_length": {{primer_length}}, "primer_gc_bounds": {{primer_gc_bounds}} } ``` ## Rules - Use the exact endpoints in `amplicon_interval`; do not move them to find better primers. - The forward primer is the first `primer_length` DNA bases of the amplicon. The reverse primer is the reverse complement of its last `primer_length` bases. Write both primers 5′→3′. - Both primers must meet the inclusive `primer_gc_bounds`. - The forward primer sequence must occur once on the supplied coding strand; the reverse primer’s reverse complement must occur once. Check uniqueness only within this CDS. - If either fixed primer fails a check, return the specified failure object rather than choosing different endpoints. ## Answer - `forward` is the amplicon’s first primer_length DNA bases. `reverse` is the reverse complement of its final primer_length bases. Write both as continuous uppercase 5′→3′ DNA strings. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "forward": <DNA string>, "reverse": <DNA string> } ``` If either fixed primer fails the supplied checks, return: ```json { "status": "impossible", "forward": null, "reverse": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `forward` is the amplicon’s first primer_length DNA bases. `reverse` is the reverse complement of its final primer_length bases. Write both as continuous uppercase 5′→3′ DNA strings. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "status": "ok", "forward": <DNA string>, "reverse": <DNA string> } If either fixed primer fails the supplied checks, return: { "status": "impossible", "forward": null, "reverse": null } Formal fields and scoring Inputs `cds` DNA string `amplicon_interval` inclusive [a,b] entirely inside CDS `primer_length` fixed integer `primer_gc_bounds` inclusive fractions Rules & boundaries Forward=first L bases of interval; reverse=reverse complement of last L. Both satisfy GC bounds and have unique oriented binding sites in this CDS. No melting-temperature or genome-wide specificity claim. Scoring `exact_match` · complete parsed answer Reference blocks DNA reverse complement, GC fraction Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "assay.primer_pair", "output_format": "json", "scorer": "exact_match", "family": "assay", "domain": "dna", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." }
Select a CDS-specific RNA probe`assay.select_rna_probe`	Choose a complementary RNA probe that matches the target CDS’s RNA but differs sufficiently from every supplied background RNA.
Select a CDS-specific RNA probe The question supplies a target CDS, comparison CDSs, a probe length, and a required minimum mismatch count against backgrounds. This defines specificity relative to a stated panel rather than the whole transcriptome. Find a target window and take its reverse complement as the probe. Reject it if any same-length background window is too close a match. Among passing probes choose the alphabetically first RNA sequence. Only exact-length, ungapped comparisons are used. Practical context RNA fluorescence in situ hybridization detects transcripts using labeled oligonucleotide probes that bind complementary RNA sequences. This task isolates the candidate-specificity screen against supplied CDS-derived RNAs. Binding strength, competing structures, and detection chemistry require additional assay design. Method references: Imaging individual mRNA molecules with labeled probes, NUPACK documentation. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `background_cds_panel`: `["ATG AAA GCA GTT TTC GTA CTG TGA"]`; `probe_length`: `12`; `background_mismatch_floor`: `1`. Procedure. Enumerate reverse complements of every 12-base target window. Reject probes having an exact complementary window in the I4V comparison, then choose the first remaining RNA string in A<C<G<U order. Final answer: AAAAUUGCUUUC Checking the answer. The probe must perfectly complement a target window and meet the mismatch minimum everywhere in every background. Report IMPOSSIBLE if no probe qualifies. Easy example · references supplied Prompt ## Task Choose an RNA probe that perfectly matches the target and meets the mismatch requirement for every comparison sequence. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "background_cds_panel": ["ATG AAA GCA GTT TTC GTA CTG TGA"], "probe_length": 12, "background_mismatch_floor": 1 } ``` ## Rules - Transcribe the target and each comparison CDS to coding RNA. - Choose a probe of exactly `probe_length` bases that is the reverse complement of at least one target window. - Compare the probe’s reverse complement with every equal-length window of every comparison RNA. Every such comparison must differ at at least `background_mismatch_floor` positions. - Use ungapped comparisons and exclude wobble pairing. ## Selection Choose the lexicographically smallest qualifying probe sequence in RNA order A<C<G<U. A probe occurring at several target positions remains one sequence choice. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNA reverse complement RNA complementary bases are A↔U and C↔G. Complement each base, then reverse the sequence to write its complementary partner 5′→3′. ## Answer - Return one continuous uppercase 5′→3′ RNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: AAAAUUGCUUUC` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Comparison is a designed I4V minigene variant (ATT→GTT). Probes and escape changes are computed design exercises, not validated reagents. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Choose an RNA probe that perfectly matches the target and meets the mismatch requirement for every comparison sequence. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "background_cds_panel": {{background_cds_panel}}, "probe_length": {{probe_length}}, "background_mismatch_floor": {{background_mismatch_floor}} } ``` ## Rules - Transcribe the target and each comparison CDS to coding RNA. - Choose a probe of exactly `probe_length` bases that is the reverse complement of at least one target window. - Compare the probe’s reverse complement with every equal-length window of every comparison RNA. Every such comparison must differ at at least `background_mismatch_floor` positions. - Use ungapped comparisons and exclude wobble pairing. ## Selection Choose the lexicographically smallest qualifying probe sequence in RNA order A<C<G<U. A probe occurring at several target positions remains one sequence choice. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNA reverse complement RNA complementary bases are A↔U and C↔G. Complement each base, then reverse the sequence to write its complementary partner 5′→3′. ## Answer - Return one continuous uppercase 5′→3′ RNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Choose an RNA probe that perfectly matches the target and meets the mismatch requirement for every comparison sequence. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "background_cds_panel": ["ATG AAA GCA GTT TTC GTA CTG TGA"], "probe_length": 12, "background_mismatch_floor": 1 } ``` ## Rules - Transcribe the target and each comparison CDS to coding RNA. - Choose a probe of exactly `probe_length` bases that is the reverse complement of at least one target window. - Compare the probe’s reverse complement with every equal-length window of every comparison RNA. Every such comparison must differ at at least `background_mismatch_floor` positions. - Use ungapped comparisons and exclude wobble pairing. ## Selection Choose the lexicographically smallest qualifying probe sequence in RNA order A<C<G<U. A probe occurring at several target positions remains one sequence choice. ## Answer - Return one continuous uppercase 5′→3′ RNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: AAAAUUGCUUUC` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Comparison is a designed I4V minigene variant (ATT→GTT). Probes and escape changes are computed design exercises, not validated reagents. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Choose an RNA probe that perfectly matches the target and meets the mismatch requirement for every comparison sequence. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "background_cds_panel": {{background_cds_panel}}, "probe_length": {{probe_length}}, "background_mismatch_floor": {{background_mismatch_floor}} } ``` ## Rules - Transcribe the target and each comparison CDS to coding RNA. - Choose a probe of exactly `probe_length` bases that is the reverse complement of at least one target window. - Compare the probe’s reverse complement with every equal-length window of every comparison RNA. Every such comparison must differ at at least `background_mismatch_floor` positions. - Use ungapped comparisons and exclude wobble pairing. ## Selection Choose the lexicographically smallest qualifying probe sequence in RNA order A<C<G<U. A probe occurring at several target positions remains one sequence choice. ## Answer - Return one continuous uppercase 5′→3′ RNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase 5′→3′ RNA sequence. Selection Choose the lexicographically smallest qualifying probe sequence in RNA order A<C<G<U. A probe occurring at several target positions remains one sequence choice. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `rna` target coding RNA `background_rnas` coding RNAs derived from supplied comparison CDSs `probe_length` fixed integer `background_mismatch_floor` positive integer Rules & boundaries Perfect target complement; every same-length background window has at least floor mismatches. No gaps/wobble. Pick lexicographically smallest valid RNA probe. Choose the lexicographically smallest qualifying probe sequence in RNA order A<C<G<U. A probe occurring at several target positions remains one sequence choice. Scoring `exact_match` · complete parsed answer Reference blocks Coding DNA to RNA, RNA reverse complement Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "assay.select_rna_probe", "output_format": "scalar", "scorer": "exact_match", "family": "assay", "domain": "rna", "objective": "none", "selection_rule": "Choose the lexicographically smallest qualifying probe sequence in RNA order A<C<G<U. A probe occurring at several target positions remains one sequence choice." } Used in: Choose an RNA probe and a matched escape control
Distinguish CDSs with a probe panel`assay.min_probe_panel`	Select the smallest subset of supplied RNA probes whose match patterns distinguish every CDS in a supplied collection.
Distinguish CDSs with a probe panel Each candidate probe gives a binary measurement: it either has an exact complementary match in a CDS-derived RNA or it does not. Several probes together give each RNA a pattern of positive and negative results. Choose probes so no two RNAs have the same resulting pattern. Minimize the number chosen; use the specified probe-ID ordering to break ties. This is a measurement-selection problem, not a request to predict hybridization efficiency. Practical context Choosing a small set of informative probes can reduce the number of measurements needed to distinguish candidate RNAs. The sequence-to-panel composition supplies the match matrix. This is the combinatorial selection stage, using ideal binary hits rather than measured hybridization performance. Task Outline Inputs. `cds_panel`: `["ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ATG AAA GCC ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGT ACT TCC TGA"]`; `candidate_probes`: supplied table shown in the prompt. Procedure. For every candidate probe, record exact complementary hits in the three CDS-derived RNAs. Enumerate probe subsets by size, require three different hit signatures, and apply the probe-ID tie-break. Final answer: ["trpL_A3_window", "trpL_R12_window"] Checking the answer. Every RNA must have a unique signature and the panel must use the minimum possible number of probes. Easy example · references supplied Prompt ## Task Choose the smallest set of candidate RNA probes whose match patterns distinguish every CDS in the panel. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_panel": [ "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ATG AAA GCC ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGT ACT TCC TGA" ], "candidate_probes": { "trpL_A3_window": "GAAAAUUGCUUU", "trpL_R12_window": "AGUGCGCCACCA", "trpL_common": "ACCUUUCAGUAC" } } ``` ## Rules - Transcribe each CDS to coding RNA and use only the supplied candidate probes. - A probe hits an RNA if at least one equal-length window is its exact reverse complement. - For each RNA, record one hit/no-hit bit per selected probe in the selected probe order. Every input RNA must have a different bit string. ## Selection Minimize the number of selected probes, then choose the lexicographically smallest sorted list of probe IDs. Compare complete ID strings in ASCII order. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNA reverse complement RNA complementary bases are A↔U and C↔G. Complement each base, then reverse the sequence to write its complementary partner 5′→3′. ## Answer - Return a list of double-quoted probe IDs, spelled exactly as supplied, unique and sorted in ASCII order. - A valid zero-probe panel is []. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ["trpL_A3_window", "trpL_R12_window"]` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Panel order: native trpL, synonymous Ala3 GCA→GCC variant, synonymous Arg12 CGC→CGT variant. Candidate 12-base probes are designed reverse complements of native windows 4–15, 28–39, and 16–27. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Choose the smallest set of candidate RNA probes whose match patterns distinguish every CDS in the panel. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_panel": {{cds_panel}}, "candidate_probes": {{candidate_probes}} } ``` ## Rules - Transcribe each CDS to coding RNA and use only the supplied candidate probes. - A probe hits an RNA if at least one equal-length window is its exact reverse complement. - For each RNA, record one hit/no-hit bit per selected probe in the selected probe order. Every input RNA must have a different bit string. ## Selection Minimize the number of selected probes, then choose the lexicographically smallest sorted list of probe IDs. Compare complete ID strings in ASCII order. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNA reverse complement RNA complementary bases are A↔U and C↔G. Complement each base, then reverse the sequence to write its complementary partner 5′→3′. ## Answer - Return a list of double-quoted probe IDs, spelled exactly as supplied, unique and sorted in ASCII order. - A valid zero-probe panel is []. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Choose the smallest set of candidate RNA probes whose match patterns distinguish every CDS in the panel. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_panel": [ "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ATG AAA GCC ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGT ACT TCC TGA" ], "candidate_probes": { "trpL_A3_window": "GAAAAUUGCUUU", "trpL_R12_window": "AGUGCGCCACCA", "trpL_common": "ACCUUUCAGUAC" } } ``` ## Rules - Transcribe each CDS to coding RNA and use only the supplied candidate probes. - A probe hits an RNA if at least one equal-length window is its exact reverse complement. - For each RNA, record one hit/no-hit bit per selected probe in the selected probe order. Every input RNA must have a different bit string. ## Selection Minimize the number of selected probes, then choose the lexicographically smallest sorted list of probe IDs. Compare complete ID strings in ASCII order. ## Answer - Return a list of double-quoted probe IDs, spelled exactly as supplied, unique and sorted in ASCII order. - A valid zero-probe panel is []. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ["trpL_A3_window", "trpL_R12_window"]` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Panel order: native trpL, synonymous Ala3 GCA→GCC variant, synonymous Arg12 CGC→CGT variant. Candidate 12-base probes are designed reverse complements of native windows 4–15, 28–39, and 16–27. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Choose the smallest set of candidate RNA probes whose match patterns distinguish every CDS in the panel. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_panel": {{cds_panel}}, "candidate_probes": {{candidate_probes}} } ``` ## Rules - Transcribe each CDS to coding RNA and use only the supplied candidate probes. - A probe hits an RNA if at least one equal-length window is its exact reverse complement. - For each RNA, record one hit/no-hit bit per selected probe in the selected probe order. Every input RNA must have a different bit string. ## Selection Minimize the number of selected probes, then choose the lexicographically smallest sorted list of probe IDs. Compare complete ID strings in ASCII order. ## Answer - Return a list of double-quoted probe IDs, spelled exactly as supplied, unique and sorted in ASCII order. - A valid zero-probe panel is []. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return a list of double-quoted probe IDs, spelled exactly as supplied, unique and sorted in ASCII order. A valid zero-probe panel is []. Selection Minimize the number of selected probes, then choose the lexicographically smallest sorted list of probe IDs. Compare complete ID strings in ASCII order. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `coding_rnas` coding RNAs derived from a supplied CDS panel `candidate_probes` named 5′→3′ RNA probes Rules & boundaries Binary hit=any exact reverse-complement match. Chosen probes give every RNA a unique binary signature. Minimize count; tie-break sorted ID list lexicographically. Minimize the number of selected probes, then choose the lexicographically smallest sorted list of probe IDs. Compare complete ID strings in ASCII order. Design objective minimize number of probes Scoring `exact_match` · complete parsed answer Reference blocks Coding DNA to RNA, RNA reverse complement Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "assay.min_probe_panel", "output_format": "scalar", "scorer": "exact_match", "family": "assay", "domain": "rna", "objective": "minimize number of probes", "selection_rule": "Minimize the number of selected probes, then choose the lexicographically smallest sorted list of probe IDs. Compare complete ID strings in ASCII order." } Used in: From CDS collection to a minimal measurement panel

09 /

Overlapping reading frames

Explicitly annotated overlaps and coupled constraints.

2 tasks

overlap task definitions
Task	Description	Expand details
Evaluate a change in two frames`overlap.compare_products`	Given two annotated coding windows and a DNA substitution, compare the before-and-after peptide in each window.
Evaluate a change in two frames Overlapping coding information occurs naturally. In the human CDKN2A locus, p16INK4a and p14ARF use different reading frames in a shared exon. One DNA edit can therefore affect two encoded products. Here both coding intervals and strands are supplied; no gene or splice-site discovery is required. Apply the equal-length DNA replacement and translate each annotated window before and after it. For these windows, use internal codon meanings throughout and retain stops as * rather than truncating: the goal is to expose every codon-level consequence. Role in related tasks When two annotated coding windows overlap, a DNA edit can change both products. The overlapping-product audit uses this comparison to check the intended edit and the product that must remain fixed. The windows are supplied; discovering overlapping genes is a different problem. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `frames`: `[{"interval": [7, 12], "strand": "+"}, {"interval": [6, 11], "strand": "+"}]`; `replacement`: `[8, 8, "T"]`. Procedure. Translate window A (7–12) as AI and window B (6–11) as SN using internal-codon meanings. C8T changes A’s GCA to GTA but changes B’s AGC to synonymous AGT; report both before/after products. Final answer: [ { "frame": "A", "original_peptide": "AI", "edited_peptide": "VI", "changed": true }, { "frame": "B", "original_peptide": "SN", "edited_peptide": "SN", "changed": false } ] Checking the answer. All four peptide strings and the changed/unchanged flags must agree with the annotated windows and replacement. Easy example · references supplied Prompt ## Task Report how the supplied DNA replacement changes the peptides encoded by the two annotated reading windows. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `replacement` is `[first_base, last_base, new_text]` in the original DNA coordinates. `frames` lists frame A first and frame B second. Each record supplies its inclusive DNA interval and strand. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "frames": [ { "interval": [7, 12], "strand": "+" }, { "interval": [6, 11], "strand": "+" } ], "replacement": [8, 8, "T"] } ``` ## Rules - For each supplied window, extract its full DNA interval and reverse-complement it if its strand is `-`. - Translate every complete triplet under ordinary internal NCBI genetic code 11 meanings. Do not treat the first triplet as an initiator. - Keep stop symbols as `` and continue translating subsequent triplets. - Apply the equal-length DNA replacement to the original CDS coordinates, then repeat the same window extraction and translation. ## Reference material ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return exactly two records in supplied frame order: "A" for the first annotated window and "B" for the second. - `original_peptide` and `edited_peptide` are the full internal translations before and after editing, as continuous uppercase strings retaining every stop as . `changed` is true exactly when the two strings differ. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "frame": <"A" or "B">, "original_peptide": <amino-acid string, retaining stops as >, "edited_peptide": <amino-acid string, retaining stops as >, "changed": <boolean> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: [ { "frame": "A", "original_peptide": "AI", "edited_peptide": "VI", "changed": true }, { "frame": "B", "original_peptide": "SN", "edited_peptide": "SN", "changed": false } ]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. These are deliberately annotated overlapping test windows on native trpL, not a claim that the second frame is a natural gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Report how the supplied DNA replacement changes the peptides encoded by the two annotated reading windows. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `replacement` is `[first_base, last_base, new_text]` in the original DNA coordinates. `frames` lists frame A first and frame B second. Each record supplies its inclusive DNA interval and strand. ```json { "cds": "{{cds}}", "frames": {{frames}}, "replacement": {{replacement}} } ``` ## Rules - For each supplied window, extract its full DNA interval and reverse-complement it if its strand is `-`. - Translate every complete triplet under ordinary internal NCBI genetic code 11 meanings. Do not treat the first triplet as an initiator. - Keep stop symbols as `` and continue translating subsequent triplets. - Apply the equal-length DNA replacement to the original CDS coordinates, then repeat the same window extraction and translation. ## Reference material ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return exactly two records in supplied frame order: "A" for the first annotated window and "B" for the second. - `original_peptide` and `edited_peptide` are the full internal translations before and after editing, as continuous uppercase strings retaining every stop as . `changed` is true exactly when the two strings differ. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "frame": <"A" or "B">, "original_peptide": <amino-acid string, retaining stops as >, "edited_peptide": <amino-acid string, retaining stops as >, "changed": <boolean> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Report how the supplied DNA replacement changes the peptides encoded by the two annotated reading windows. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `replacement` is `[first_base, last_base, new_text]` in the original DNA coordinates. `frames` lists frame A first and frame B second. Each record supplies its inclusive DNA interval and strand. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "frames": [ { "interval": [7, 12], "strand": "+" }, { "interval": [6, 11], "strand": "+" } ], "replacement": [8, 8, "T"] } ``` ## Rules - For each supplied window, extract its full DNA interval and reverse-complement it if its strand is `-`. - Translate every complete triplet under ordinary internal NCBI genetic code 11 meanings. Do not treat the first triplet as an initiator. - Keep stop symbols as `` and continue translating subsequent triplets. - Apply the equal-length DNA replacement to the original CDS coordinates, then repeat the same window extraction and translation. ## Answer - Return exactly two records in supplied frame order: "A" for the first annotated window and "B" for the second. - `original_peptide` and `edited_peptide` are the full internal translations before and after editing, as continuous uppercase strings retaining every stop as . `changed` is true exactly when the two strings differ. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "frame": <"A" or "B">, "original_peptide": <amino-acid string, retaining stops as >, "edited_peptide": <amino-acid string, retaining stops as >, "changed": <boolean> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: [ { "frame": "A", "original_peptide": "AI", "edited_peptide": "VI", "changed": true }, { "frame": "B", "original_peptide": "SN", "edited_peptide": "SN", "changed": false } ]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. These are deliberately annotated overlapping test windows on native trpL, not a claim that the second frame is a natural gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Report how the supplied DNA replacement changes the peptides encoded by the two annotated reading windows. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `replacement` is `[first_base, last_base, new_text]` in the original DNA coordinates. `frames` lists frame A first and frame B second. Each record supplies its inclusive DNA interval and strand. ```json { "cds": "{{cds}}", "frames": {{frames}}, "replacement": {{replacement}} } ``` ## Rules - For each supplied window, extract its full DNA interval and reverse-complement it if its strand is `-`. - Translate every complete triplet under ordinary internal NCBI genetic code 11 meanings. Do not treat the first triplet as an initiator. - Keep stop symbols as `` and continue translating subsequent triplets. - Apply the equal-length DNA replacement to the original CDS coordinates, then repeat the same window extraction and translation. ## Answer - Return exactly two records in supplied frame order: "A" for the first annotated window and "B" for the second. - `original_peptide` and `edited_peptide` are the full internal translations before and after editing, as continuous uppercase strings retaining every stop as . `changed` is true exactly when the two strings differ. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "frame": <"A" or "B">, "original_peptide": <amino-acid string, retaining stops as >, "edited_peptide": <amino-acid string, retaining stops as >, "changed": <boolean> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` Return exactly two records in supplied frame order: "A" for the first annotated window and "B" for the second. `original_peptide` and `edited_peptide` are the full internal translations before and after editing, as continuous uppercase strings retaining every stop as . `changed` is true exactly when the two strings differ. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. [ { "frame": <"A" or "B">, "original_peptide": <amino-acid string, retaining stops as >, "edited_peptide": <amino-acid string, retaining stops as >, "changed": <boolean> } ] Formal fields and scoring Inputs `cds` DNA coding sequence containing both annotated intervals `frames` two records: interval [a,b], strand +\|−; complete codons, no initiation assumption `replacement` equal-length original-coordinate DNA replacement Rules & boundaries Extract each interval, reverse-complement minus strand, translate all triplets internally; represent stops as without truncating. This component is an annotated coding-window analysis, not ORF discovery. Scoring `exact_match` · complete parsed answer Reference blocks DNA reverse complement, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "overlap.compare_products", "output_format": "json", "scorer": "exact_match", "family": "overlap", "domain": "multi", "objective": "none", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Used in: Edit one frame and audit both products
Change one frame, preserve the other`overlap.constrained_edit`	Change DNA so one annotated coding window encodes a requested peptide while the overlapping window’s peptide stays unchanged.
Change one frame, preserve the other The two coding windows and their strands are supplied, together with an exact target peptide for the first and a list of DNA positions allowed to change. The original second-window peptide is the product that must be preserved. Search substitutions at those shared DNA positions and evaluate both translations after every candidate choice. A synonymous change in one window may alter the other, so separately choosing codons for the two peptides is insufficient. Use the same internal-translation convention as the overlap-comparison task. Practical context The overlap-comparison task describes a real shared-exon example. When one DNA segment encodes two products, a mutation intended for one can recode the other. This task searches exact peptide-preserving constraints, and the verification composition reports both products so the intended protection can be checked. Method reference: NCBI CDKN2A annotation. Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `frames`: `[{"interval": [7, 12], "strand": "+"}, {"interval": [6, 11], "strand": "+"}]`; `target_peptide_a`: `VI`; `editable_positions`: `[8]`. Procedure. Test the four possible bases at editable position 8. C8T yields VI in frame A and keeps SN in frame B. Every other DNA position remains fixed. Final answer: ATGAAAGTAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA Checking the answer. One final DNA sequence must satisfy both peptide requirements and the edit permissions. Return IMPOSSIBLE when the coupled constraints have no solution. Easy example · references supplied Prompt ## Task Change the peptide in annotated frame A to `target_peptide_a` while preserving the peptide in frame B. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `frames` lists frame A first and frame B second. Each record supplies its inclusive DNA interval and strand. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "frames": [ { "interval": [7, 12], "strand": "+" }, { "interval": [6, 11], "strand": "+" } ], "target_peptide_a": "VI", "editable_positions": [8] } ``` ## Rules - Only DNA positions in `editable_positions` may change, using substitutions without changing the sequence length. - Extract each annotated DNA window, reversing and complementing minus-strand windows. Use ordinary internal NCBI genetic code 11 meanings throughout, without an initiation override. - Keep stops as `` and continue past them. - The final A peptide must equal `target_peptide_a`; the final B peptide must equal its original complete peptide. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGTAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA` Native CDS: E. coli* K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The two overlapping test windows are supplied annotations, not two natural genes. C8T changes frame A Ala→Val while preserving the overlapping frame B Ser codon AGC→AGT. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Change the peptide in annotated frame A to `target_peptide_a` while preserving the peptide in frame B. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `frames` lists frame A first and frame B second. Each record supplies its inclusive DNA interval and strand. ```json { "cds": "{{cds}}", "frames": {{frames}}, "target_peptide_a": "{{target_peptide_a}}", "editable_positions": {{editable_positions}} } ``` ## Rules - Only DNA positions in `editable_positions` may change, using substitutions without changing the sequence length. - Extract each annotated DNA window, reversing and complementing minus-strand windows. Use ordinary internal NCBI genetic code 11 meanings throughout, without an initiation override. - Keep stops as `` and continue past them. - The final A peptide must equal `target_peptide_a`; the final B peptide must equal its original complete peptide. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Change the peptide in annotated frame A to `target_peptide_a` while preserving the peptide in frame B. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `frames` lists frame A first and frame B second. Each record supplies its inclusive DNA interval and strand. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "frames": [ { "interval": [7, 12], "strand": "+" }, { "interval": [6, 11], "strand": "+" } ], "target_peptide_a": "VI", "editable_positions": [8] } ``` ## Rules - Only DNA positions in `editable_positions` may change, using substitutions without changing the sequence length. - Extract each annotated DNA window, reversing and complementing minus-strand windows. Use ordinary internal NCBI genetic code 11 meanings throughout, without an initiation override. - Keep stops as `` and continue past them. - The final A peptide must equal `target_peptide_a`; the final B peptide must equal its original complete peptide. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGTAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The two overlapping test windows are supplied annotations, not two natural genes. C8T changes frame A Ala→Val while preserving the overlapping frame B Ser codon AGC→AGT. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Change the peptide in annotated frame A to `target_peptide_a` while preserving the peptide in frame B. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `frames` lists frame A first and frame B second. Each record supplies its inclusive DNA interval and strand. ```json { "cds": "{{cds}}", "frames": {{frames}}, "target_peptide_a": "{{target_peptide_a}}", "editable_positions": {{editable_positions}} } ``` ## Rules - Only DNA positions in `editable_positions` may change, using substitutions without changing the sequence length. - Extract each annotated DNA window, reversing and complementing minus-strand windows. Use ordinary internal NCBI genetic code 11 meanings throughout, without an initiation override. - Keep stops as `` and continue past them. - The final A peptide must equal `target_peptide_a`; the final B peptide must equal its original complete peptide. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Formal fields and scoring Inputs `cds` DNA sequence containing supplied coding windows `frames` two annotated interval/strand records `target_peptide_a` exact requested peptide in frame A `editable_positions` permitted DNA positions Rules & boundaries Frame A must equal target_peptide_a; frame B peptide must remain exactly original. Internal codon mapping, retain ; no start discovery. All unlisted bases fixed; finite small search. Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Scoring `exact_match` · complete parsed answer Reference blocks DNA reverse complement, NCBI genetic code 11: internal DNA codons Prompt templates start from a CDS and include any required transcription or translation. Literal task fields { "task_type": "overlap.constrained_edit", "output_format": "scalar", "scorer": "exact_match", "family": "overlap", "domain": "multi", "objective": "none", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." } Used in: Edit one frame and audit both products

Composition tasks (19)

Each composition task lists the tasks it combines and explains how one task’s output is used by another. Some run tasks in sequence, some use the same result for several calculations, and others require one sequence to satisfy several conditions at once.

Composition types

Pipeline: CDS → protein → peptide intervals → DNA intervals.
Branch: one protein → mass, charge, and composition.
Joint: one CDS must preserve protein, remove sites, and meet GC bounds simultaneously.

For a joint design, completing one rewrite and then another can undo earlier constraints. Every condition must be checked on the same final sequence.

Select a row to read the sequence of operations, an example, and the final success conditions. The component-task links open the corresponding task explanations. Each composition also includes paired prompt templates.

Composition tasks
Task	Description	Expand details
From peptide motif to encoding DNA`pipeline.motif_to_dna`	Find a supplied pattern in the CDS’s encoded protein, then report the DNA intervals encoding those protein matches.
From peptide motif to encoding DNA The input is a CDS plus an amino-acid pattern. First translate the CDS and find all matching residue intervals. For each match, map that interval back to the original DNA using the known reading frame. This asks the model to search at one biological level and report locations at another, rather than searching for the protein letters in DNA. Practical context This prepares a protein-level feature for DNA engineering: locate the supplied motif and identify the codons that encode it. Those coordinates can feed permitted motif removal or a planned substitution, without asking the model to infer which sequence features are functional. Component tasks Translate DNA to protein`cds.translate`Find a supplied peptide pattern`protein.motif_positions`Map residues back to DNA`coords.aa_to_nt` Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `pattern`: `WWX[ST]`. Procedure. Translate native trpL and find WWRT matching WWX[ST] at residues 10–13. Convert the interval to original DNA using 3×10−2 = 28 and 3×13 = 39. Final answer: [ { "aa_interval": [10, 13], "nt_interval": [28, 39] } ] Checking the answer. Every protein match and its corresponding DNA interval must be correct; overlapping matches remain separate results. Easy example · references supplied Prompt ## Task Find every peptide-pattern match encoded by this CDS and return both its residue interval and its encoding DNA interval. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "pattern": "WWX[ST]" } ``` ## Rules - Translate with NCBI genetic code 11, stopping at the first in-frame stop. - Pattern syntax: a letter matches itself; `X` matches any residue; `[set]` matches a listed residue; `{set}` excludes listed residues. - Include overlapping peptide-pattern matches. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - `aa_interval` is [first_residue, last_residue] in the translated protein. `nt_interval` is [first_base, last_base] in the original CDS. Both have 1-based inclusive integer endpoints. - Include every overlapping match, sorted by amino-acid start and then end. An empty result is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "aa_interval": [<first residue>, <last residue>], "nt_interval": [<first base>, <last base>] } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: [ { "aa_interval": [10, 13], "nt_interval": [28, 39] } ]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find every peptide-pattern match encoded by this CDS and return both its residue interval and its encoding DNA interval. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pattern": "{{pattern}}" } ``` ## Rules - Translate with NCBI genetic code 11, stopping at the first in-frame stop. - Pattern syntax: a letter matches itself; `X` matches any residue; `[set]` matches a listed residue; `{set}` excludes listed residues. - Include overlapping peptide-pattern matches. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - `aa_interval` is [first_residue, last_residue] in the translated protein. `nt_interval` is [first_base, last_base] in the original CDS. Both have 1-based inclusive integer endpoints. - Include every overlapping match, sorted by amino-acid start and then end. An empty result is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "aa_interval": [<first residue>, <last residue>], "nt_interval": [<first base>, <last base>] } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Find every peptide-pattern match encoded by this CDS and return both its residue interval and its encoding DNA interval. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "pattern": "WWX[ST]" } ``` ## Rules - Translate with NCBI genetic code 11, stopping at the first in-frame stop. - Pattern syntax: a letter matches itself; `X` matches any residue; `[set]` matches a listed residue; `{set}` excludes listed residues. - Include overlapping peptide-pattern matches. ## Answer - `aa_interval` is [first_residue, last_residue] in the translated protein. `nt_interval` is [first_base, last_base] in the original CDS. Both have 1-based inclusive integer endpoints. - Include every overlapping match, sorted by amino-acid start and then end. An empty result is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "aa_interval": [<first residue>, <last residue>], "nt_interval": [<first base>, <last base>] } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: [ { "aa_interval": [10, 13], "nt_interval": [28, 39] } ]` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find every peptide-pattern match encoded by this CDS and return both its residue interval and its encoding DNA interval. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pattern": "{{pattern}}" } ``` ## Rules - Translate with NCBI genetic code 11, stopping at the first in-frame stop. - Pattern syntax: a letter matches itself; `X` matches any residue; `[set]` matches a listed residue; `{set}` excludes listed residues. - Include overlapping peptide-pattern matches. ## Answer - `aa_interval` is [first_residue, last_residue] in the translated protein. `nt_interval` is [first_base, last_base] in the original CDS. Both have 1-based inclusive integer endpoints. - Include every overlapping match, sorted by amino-acid start and then end. An empty result is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "aa_interval": [<first residue>, <last residue>], "nt_interval": [<first base>, <last base>] } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `aa_interval` is [first_residue, last_residue] in the translated protein. `nt_interval` is [first_base, last_base] in the original CDS. Both have 1-based inclusive integer endpoints. Include every overlapping match, sorted by amino-acid start and then end. An empty result is []. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. [ { "aa_interval": [<first residue>, <last residue>], "nt_interval": [<first base>, <last base>] } ] Steps and final output Translate the CDS, then find the motif in that protein. Convert each matching residue interval into nucleotide coordinates using the original CDS. Output: JSON array of records in the specified order Literal task fields { "composition_type": "pipeline.motif_to_dna", "output_format": "json", "scorer": "exact_match", "mode": "pipeline", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Reference material. Annotated CDS initiation, NCBI genetic code 11: internal DNA codons
From CDS to defined protein measurements`pipeline.protein_properties`	Translate one CDS and calculate a fixed set of measurements from that same encoded protein.
From CDS to defined protein measurements The question supplies a CDS, a pH, and pKa values. Translate once, then use the same product for residue counts, mass, mean hydropathy, charge at the supplied pH, and pI. This is a branching composition: no measurement changes the sequence passed to another measurement. Practical context This resembles a construct-characterization report: theoretical mass supports product identification, and charge or pI supports purification planning. Hydropathy and composition add descriptive context. The measurements answer separate questions; combining them does not create a general predictor of protein expression or function. Method references: ExPASy ProtParam documentation, Cytiva ion-exchanger selection guide. Component tasks Translate DNA to protein`cds.translate`Count amino acids`protein.residue_counts`Calculate peptide mass`protein.monoisotopic_mass`Calculate mean hydropathy`protein.mean_hydropathy`Calculate model charge at a pH`protein.charge_at_ph`Calculate the isoelectric point (pI)`protein.isoelectric_point` Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `ph`: `7.0`; `pka`: supplied table shown in the prompt. Procedure. Translate once to MKAIFVLKGWWRTS. Use that same sequence for counts, residue-mass sum plus one water, mean hydropathy, charge at pH 7, and the pH where that charge becomes zero. Final answer: { "residue_counts": { "M": 1, "K": 2, "A": 1, "I": 1, "F": 1, "V": 1, "L": 1, "G": 1, "W": 2, "R": 1, "T": 1, "S": 1 }, "mass_da": 1721.944, "mean_hydropathy": 0.2143, "charge": 2.9095, "pi": 11.67 } Checking the answer. Each named measurement is checked under its own stated convention and numerical precision. Easy example · references supplied Prompt ## Task Translate the CDS and report residue counts, neutral monoisotopic mass, mean Kyte–Doolittle hydropathy, charge at ph, and pI under pka. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ph": 7.0, "pka": { "N_term": 8.0, "C_term": 4.0, "K": 10.5, "R": 12.5, "H": 6.0, "D": 3.9, "E": 4.1, "C": 8.3, "Y": 10.1 } } ``` ## Rules - Translate once with NCBI genetic code 11, stopping at the first in-frame stop. Use that same complete protein for every measurement. - Use an unmodified linear protein with both termini free. Do not remove the initiator or form disulfides. - Use the independent-site Henderson–Hasselbalch calculation with the supplied `pka` values. Search pH 0–14 for pI. ## Mass constants Use these isotope masses in Da as exact decimal constants: ```text 12C = 12.00000000000 1H = 1.00782503223 14N = 14.00307400443 16O = 15.99491461957 32S = 31.9720711744 ``` For a neutral, unmodified linear peptide, sum the neutral free-amino-acid molecular formulas and subtract one H2O for each peptide bond. Equivalently, sum the incorporated residue masses and add one H2O for the complete peptide. Do not round intermediate values. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Monoisotopic residue masses Use these incorporated residue masses in Da as exact decimal constants. Each is the corresponding neutral free amino acid minus one H2O. ```text A=71.03711378515 C=103.00918495955 D=115.02694302429 E=129.04259308875 F=147.06841391407 G=57.02146372069 H=137.05891185847 I=113.08406397853 K=128.09496301519 L=113.08406397853 M=131.04048508847 N=114.04292744138 P=97.05276384961 Q=128.05857750584 R=156.10111102405 S=87.03202840472 T=101.04767846918 V=99.06841391407 W=186.07931295073 Y=163.06332853364 ``` Add one H2O = 18.01056468403 Da for the complete free linear peptide. ### Kyte–Doolittle hydropathy scores Use these Kyte–Doolittle residue scores. Whole-protein mean hydropathy is their arithmetic mean, counting every residue occurrence. ```text A=1.8 R=-4.5 N=-3.5 D=-3.5 C=2.5 Q=-3.5 E=-3.5 G=-0.4 H=-3.2 I=4.5 L=3.8 K=-3.9 M=1.9 F=2.8 P=-1.6 S=-0.8 T=-0.7 W=-0.9 Y=-1.3 V=4.2 ``` ### Ionization and net charge A basic site contributes +1/(1+10^(pH−pKa)); an acidic site contributes −1/(1+10^(pKa−pH)). Basic sites: the N terminus and K, R, H side chains. Acidic sites: the C terminus and D, E, C, Y side chains. Count each terminus once and each listed side chain once per residue occurrence; sum their contributions. ## Answer - `residue_counts` maps each observed one-letter amino-acid symbol to its positive integer count; omit absent residues. - `mass_da` is neutral monoisotopic mass including one water. `mean_hydropathy` is the mean Kyte–Doolittle score. - `charge` is net charge at the supplied ph. `pi` is the zero-charge pH using the supplied pka values. - Use JSON numbers rounded once: mass_da, mean_hydropathy, and charge to 4 decimal places; pi to 2. Use the component tasks’ constants and rounding rules. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "residue_counts": {"<one-letter amino-acid symbol>": <positive integer count>}, "mass_da": <number>, "mean_hydropathy": <number>, "charge": <number>, "pi": <number> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "residue_counts": { "M": 1, "K": 2, "A": 1, "I": 1, "F": 1, "V": 1, "L": 1, "G": 1, "W": 2, "R": 1, "T": 1, "S": 1 }, "mass_da": 1721.944, "mean_hydropathy": 0.2143, "charge": 2.9095, "pi": 11.67 }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Translate the CDS and report residue counts, neutral monoisotopic mass, mean Kyte–Doolittle hydropathy, charge at ph, and pI under pka. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "ph": {{ph}}, "pka": {{pka}} } ``` ## Rules - Translate once with NCBI genetic code 11, stopping at the first in-frame stop. Use that same complete protein for every measurement. - Use an unmodified linear protein with both termini free. Do not remove the initiator or form disulfides. - Use the independent-site Henderson–Hasselbalch calculation with the supplied `pka` values. Search pH 0–14 for pI. ## Mass constants Use these isotope masses in Da as exact decimal constants: ```text 12C = 12.00000000000 1H = 1.00782503223 14N = 14.00307400443 16O = 15.99491461957 32S = 31.9720711744 ``` For a neutral, unmodified linear peptide, sum the neutral free-amino-acid molecular formulas and subtract one H2O for each peptide bond. Equivalently, sum the incorporated residue masses and add one H2O for the complete peptide. Do not round intermediate values. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Monoisotopic residue masses Use these incorporated residue masses in Da as exact decimal constants. Each is the corresponding neutral free amino acid minus one H2O. ```text A=71.03711378515 C=103.00918495955 D=115.02694302429 E=129.04259308875 F=147.06841391407 G=57.02146372069 H=137.05891185847 I=113.08406397853 K=128.09496301519 L=113.08406397853 M=131.04048508847 N=114.04292744138 P=97.05276384961 Q=128.05857750584 R=156.10111102405 S=87.03202840472 T=101.04767846918 V=99.06841391407 W=186.07931295073 Y=163.06332853364 ``` Add one H2O = 18.01056468403 Da for the complete free linear peptide. ### Kyte–Doolittle hydropathy scores Use these Kyte–Doolittle residue scores. Whole-protein mean hydropathy is their arithmetic mean, counting every residue occurrence. ```text A=1.8 R=-4.5 N=-3.5 D=-3.5 C=2.5 Q=-3.5 E=-3.5 G=-0.4 H=-3.2 I=4.5 L=3.8 K=-3.9 M=1.9 F=2.8 P=-1.6 S=-0.8 T=-0.7 W=-0.9 Y=-1.3 V=4.2 ``` ### Ionization and net charge A basic site contributes +1/(1+10^(pH−pKa)); an acidic site contributes −1/(1+10^(pKa−pH)). Basic sites: the N terminus and K, R, H side chains. Acidic sites: the C terminus and D, E, C, Y side chains. Count each terminus once and each listed side chain once per residue occurrence; sum their contributions. ## Answer - `residue_counts` maps each observed one-letter amino-acid symbol to its positive integer count; omit absent residues. - `mass_da` is neutral monoisotopic mass including one water. `mean_hydropathy` is the mean Kyte–Doolittle score. - `charge` is net charge at the supplied ph. `pi` is the zero-charge pH using the supplied pka values. - Use JSON numbers rounded once: mass_da, mean_hydropathy, and charge to 4 decimal places; pi to 2. Use the component tasks’ constants and rounding rules. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "residue_counts": {"<one-letter amino-acid symbol>": <positive integer count>}, "mass_da": <number>, "mean_hydropathy": <number>, "charge": <number>, "pi": <number> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Translate the CDS and report residue counts, neutral monoisotopic mass, mean Kyte–Doolittle hydropathy, charge at ph, and pI under pka. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ph": 7.0, "pka": { "N_term": 8.0, "C_term": 4.0, "K": 10.5, "R": 12.5, "H": 6.0, "D": 3.9, "E": 4.1, "C": 8.3, "Y": 10.1 } } ``` ## Rules - Translate once with NCBI genetic code 11, stopping at the first in-frame stop. Use that same complete protein for every measurement. - Use an unmodified linear protein with both termini free. Do not remove the initiator or form disulfides. - Use the independent-site Henderson–Hasselbalch calculation with the supplied `pka` values. Search pH 0–14 for pI. ## Mass constants Use these isotope masses in Da as exact decimal constants: ```text 12C = 12.00000000000 1H = 1.00782503223 14N = 14.00307400443 16O = 15.99491461957 32S = 31.9720711744 ``` For a neutral, unmodified linear peptide, sum the neutral free-amino-acid molecular formulas and subtract one H2O for each peptide bond. Equivalently, sum the incorporated residue masses and add one H2O for the complete peptide. Do not round intermediate values. ## Answer - `residue_counts` maps each observed one-letter amino-acid symbol to its positive integer count; omit absent residues. - `mass_da` is neutral monoisotopic mass including one water. `mean_hydropathy` is the mean Kyte–Doolittle score. - `charge` is net charge at the supplied ph. `pi` is the zero-charge pH using the supplied pka values. - Use JSON numbers rounded once: mass_da, mean_hydropathy, and charge to 4 decimal places; pi to 2. Use the component tasks’ constants and rounding rules. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "residue_counts": {"<one-letter amino-acid symbol>": <positive integer count>}, "mass_da": <number>, "mean_hydropathy": <number>, "charge": <number>, "pi": <number> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "residue_counts": { "M": 1, "K": 2, "A": 1, "I": 1, "F": 1, "V": 1, "L": 1, "G": 1, "W": 2, "R": 1, "T": 1, "S": 1 }, "mass_da": 1721.944, "mean_hydropathy": 0.2143, "charge": 2.9095, "pi": 11.67 }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Translate the CDS and report residue counts, neutral monoisotopic mass, mean Kyte–Doolittle hydropathy, charge at ph, and pI under pka. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "ph": {{ph}}, "pka": {{pka}} } ``` ## Rules - Translate once with NCBI genetic code 11, stopping at the first in-frame stop. Use that same complete protein for every measurement. - Use an unmodified linear protein with both termini free. Do not remove the initiator or form disulfides. - Use the independent-site Henderson–Hasselbalch calculation with the supplied `pka` values. Search pH 0–14 for pI. ## Mass constants Use these isotope masses in Da as exact decimal constants: ```text 12C = 12.00000000000 1H = 1.00782503223 14N = 14.00307400443 16O = 15.99491461957 32S = 31.9720711744 ``` For a neutral, unmodified linear peptide, sum the neutral free-amino-acid molecular formulas and subtract one H2O for each peptide bond. Equivalently, sum the incorporated residue masses and add one H2O for the complete peptide. Do not round intermediate values. ## Answer - `residue_counts` maps each observed one-letter amino-acid symbol to its positive integer count; omit absent residues. - `mass_da` is neutral monoisotopic mass including one water. `mean_hydropathy` is the mean Kyte–Doolittle score. - `charge` is net charge at the supplied ph. `pi` is the zero-charge pH using the supplied pka values. - Use JSON numbers rounded once: mass_da, mean_hydropathy, and charge to 4 decimal places; pi to 2. Use the component tasks’ constants and rounding rules. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "residue_counts": {"<one-letter amino-acid symbol>": <positive integer count>}, "mass_da": <number>, "mean_hydropathy": <number>, "charge": <number>, "pi": <number> } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `residue_counts` maps each observed one-letter amino-acid symbol to its positive integer count; omit absent residues. `mass_da` is neutral monoisotopic mass including one water. `mean_hydropathy` is the mean Kyte–Doolittle score. `charge` is net charge at the supplied ph. `pi` is the zero-charge pH using the supplied pka values. Use JSON numbers rounded once: mass_da, mean_hydropathy, and charge to 4 decimal places; pi to 2. Use the component tasks’ constants and rounding rules. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "residue_counts": {"<one-letter amino-acid symbol>": <positive integer count>}, "mass_da": <number>, "mean_hydropathy": <number>, "charge": <number>, "pi": <number> } Steps and final output Translate the CDS once and use that same protein for all five measurements. Both prompt variants supply the pH and any custom pKa values. Return all five named measurements. Output: JSON object with exactly residue_counts, mass_da, mean_hydropathy, charge, pi Literal task fields { "composition_type": "pipeline.protein_properties", "output_format": "json", "scorer": "exact_match", "mode": "branch", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Reference material. Annotated CDS initiation, NCBI genetic code 11: internal DNA codons, Monoisotopic isotope masses, Monoisotopic residue masses, Kyte–Doolittle hydropathy scores, Ionization and net charge
From CDS panel to a diagnostic peptide`pipeline.diagnostic_digest`	From target and comparison CDSs, identify a protein-digest peptide that distinguishes the target from the supplied panel.
From CDS panel to a diagnostic peptide Start with one target CDS encoding the protein you want to identify, and comparison CDSs encoding the other proteins you want to distinguish it from. The comparison set is called the background in the input fields. Translate every CDS and digest every protein using the same trypsin rule. Discard target peptides also found in any comparison digest, then apply the peptide-length limits. Choose the shortest remaining peptide, breaking ties alphabetically; return NONE if none remains. The result distinguishes the target only from the supplied comparison set. Practical context This is a first stage in choosing a proteomics readout that distinguishes related proteins. Deriving the peptide from each actual CDS checks that its uniqueness survives translation and digestion. Experimental assay development must then establish that the candidate peptide is observable and measurable. Method reference: ExPASy PeptideMass documentation. Component tasks Translate DNA to protein`cds.translate`Digest with trypsin rules`protein.trypsin_digest`Select a distinguishing peptide`protein.diagnostic_peptide` Task Outline Inputs. `target_cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `background_cds_panel`: `["ATG AAA GCA GTT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA"]`; `length_bounds`: `[4, 10]`. Procedure. Translate and digest both the native target and I4V comparison. Their shared fragments are MK, GWWR, and TS; AIFVLK appears only in the target digest and meets the length bounds. Final answer: AIFVLK Checking the answer. The final peptide must be eligible and correctly selected from the actual translated-and-digested products, or NONE must be justified by the absence of eligible peptides. Easy example · references supplied Prompt ## Task Choose a tryptic peptide that distinguishes the target CDS’s protein from every protein encoded by the comparison panel. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "target_cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "background_cds_panel": ["ATG AAA GCA GTT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA"], "length_bounds": [4, 10] } ``` ## Rules - Translate every CDS under NCBI genetic code 11, ending each protein at its first in-frame stop. - Digest every protein completely with trypsin, with zero missed cleavage sites. - A peptide qualifies only if its sequence occurs in the target digest and in none of the comparison digests. Ignore how many copies occur. - Its length must satisfy both inclusive `length_bounds`. ## Selection From eligible target-only peptide sequences, choose the shortest, then the ASCII-lexicographically smallest sequence. Multiple copies of one sequence do not create separate choices. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Trypsin cleavage specificity Trypsin cleaves after K or R, except when the next residue is P. The K or R remains on the upstream peptide. ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. If no target-only peptide meets the length limits, return: ```text NONE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: AIFVLK` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL is the target. The comparison is a designed I4V variant (codon 4 ATT→GTT), not a second natural accession. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Choose a tryptic peptide that distinguishes the target CDS’s protein from every protein encoded by the comparison panel. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "target_cds": "{{target_cds}}", "background_cds_panel": {{background_cds_panel}}, "length_bounds": {{length_bounds}} } ``` ## Rules - Translate every CDS under NCBI genetic code 11, ending each protein at its first in-frame stop. - Digest every protein completely with trypsin, with zero missed cleavage sites. - A peptide qualifies only if its sequence occurs in the target digest and in none of the comparison digests. Ignore how many copies occur. - Its length must satisfy both inclusive `length_bounds`. ## Selection From eligible target-only peptide sequences, choose the shortest, then the ASCII-lexicographically smallest sequence. Multiple copies of one sequence do not create separate choices. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Trypsin cleavage specificity Trypsin cleaves after K or R, except when the next residue is P. The K or R remains on the upstream peptide. ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. If no target-only peptide meets the length limits, return: ```text NONE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Choose a tryptic peptide that distinguishes the target CDS’s protein from every protein encoded by the comparison panel. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "target_cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "background_cds_panel": ["ATG AAA GCA GTT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA"], "length_bounds": [4, 10] } ``` ## Rules - Translate every CDS under NCBI genetic code 11, ending each protein at its first in-frame stop. - Digest every protein completely with trypsin, with zero missed cleavage sites. - A peptide qualifies only if its sequence occurs in the target digest and in none of the comparison digests. Ignore how many copies occur. - Its length must satisfy both inclusive `length_bounds`. ## Selection From eligible target-only peptide sequences, choose the shortest, then the ASCII-lexicographically smallest sequence. Multiple copies of one sequence do not create separate choices. ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. If no target-only peptide meets the length limits, return: ```text NONE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: AIFVLK` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL is the target. The comparison is a designed I4V variant (codon 4 ATT→GTT), not a second natural accession. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Choose a tryptic peptide that distinguishes the target CDS’s protein from every protein encoded by the comparison panel. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "target_cds": "{{target_cds}}", "background_cds_panel": {{background_cds_panel}}, "length_bounds": {{length_bounds}} } ``` ## Rules - Translate every CDS under NCBI genetic code 11, ending each protein at its first in-frame stop. - Digest every protein completely with trypsin, with zero missed cleavage sites. - A peptide qualifies only if its sequence occurs in the target digest and in none of the comparison digests. Ignore how many copies occur. - Its length must satisfy both inclusive `length_bounds`. ## Selection From eligible target-only peptide sequences, choose the shortest, then the ASCII-lexicographically smallest sequence. Multiple copies of one sequence do not create separate choices. ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. If no target-only peptide meets the length limits, return: ```text NONE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. Selection From eligible target-only peptide sequences, choose the shortest, then the ASCII-lexicographically smallest sequence. Multiple copies of one sequence do not create separate choices. If no target-only peptide meets the length limits, return: NONE Steps and final output Translate and digest the target CDS and each comparison CDS (the background set). Compare the target peptides with all comparison peptides. Select the shortest target-only peptide, breaking ties alphabetically, or return NONE if no peptide meets the length limits. Output: one peptide or NONE Literal task fields { "composition_type": "pipeline.diagnostic_digest", "output_format": "scalar", "scorer": "exact_match", "mode": "pipeline", "selection_rule": "From eligible target-only peptide sequences, choose the shortest, then the ASCII-lexicographically smallest sequence. Multiple copies of one sequence do not create separate choices." } Reference material. Annotated CDS initiation, NCBI genetic code 11: internal DNA codons, Trypsin cleavage specificity
From DNA edits to changed peptide readouts`pipeline.variant_peptides`	Apply DNA edits and determine which tryptic peptide sequences are lost or gained relative to the original CDS.
From DNA edits to changed peptide readouts Keep two branches: the original CDS and a copy with the supplied DNA edits. Translate and digest both. Compare their peptide-sequence sets to identify readouts removed or introduced by the edit. An edit can change not only a peptide’s letters but also a cleavage site and therefore the fragment boundaries. Practical context An engineered DNA change can alter a peptide sequence or create or remove a cleavage boundary. Following both branches identifies candidate readouts for checking whether the edited protein product was made, before developing a mass-spectrometry assay around those products. Method reference: ExPASy PeptideMass documentation. Component tasks Apply a DNA edit script`variant.apply`Translate DNA to protein`cds.translate`Digest with trypsin rules`protein.trypsin_digest` Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `edits`: `[{"replace": [4, 6, "GCA"]}]`. Procedure. The K2A DNA change removes the first trypsin cut. Original MK and AIFVLK merge to MAAIFVLK; compare peptide sequence sets to report lost and gained readouts. Final answer: { "lost": ["AIFVLK", "MK"], "gained": ["MAAIFVLK"] } Checking the answer. Both lost and gained sets must match the two digests. This composition ignores peptide multiplicity when taking those set differences. Easy example · references supplied Prompt ## Task Apply the DNA edits and report which complete tryptic peptide sequences are lost and gained relative to the original CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "edits": [ { "replace": [4, 6, "GCA"] } ] } ``` ## Rules - Use NCBI genetic code 11 with first-stop termination. - Apply the supplied edits simultaneously to the original DNA coordinates. - Translate and completely digest both versions under `trypsin_KR_notP_v1`, with zero missed cleavages. - Compare peptide sequence sets, ignoring copy number. ## Reference material ### Initiation after DNA editing At the first DNA codon, TTG, CTG, ATT, ATC, ATA, ATG, and GTG encode initiating M under NCBI genetic code 11. If the first edited codon is not one of these initiators, use its ordinary internal assignment. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Trypsin cleavage specificity Trypsin cleaves after K or R, except when the next residue is P. The K or R remains on the upstream peptide. ## Answer - `lost` lists distinct peptide sequences present in the original tryptic digest and absent from the edited digest. `gained` lists those present only in the edited digest. - Sort each list in ASCII order. Each sequence appears once; an empty difference is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "lost": [<amino-acid string>], "gained": [<amino-acid string>] } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "lost": ["AIFVLK", "MK"], "gained": ["MAAIFVLK"] }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Apply the DNA edits and report which complete tryptic peptide sequences are lost and gained relative to the original CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "edits": {{edits}} } ``` ## Rules - Use NCBI genetic code 11 with first-stop termination. - Apply the supplied edits simultaneously to the original DNA coordinates. - Translate and completely digest both versions under `trypsin_KR_notP_v1`, with zero missed cleavages. - Compare peptide sequence sets, ignoring copy number. ## Reference material ### Initiation after DNA editing At the first DNA codon, TTG, CTG, ATT, ATC, ATA, ATG, and GTG encode initiating M under NCBI genetic code 11. If the first edited codon is not one of these initiators, use its ordinary internal assignment. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Trypsin cleavage specificity Trypsin cleaves after K or R, except when the next residue is P. The K or R remains on the upstream peptide. ## Answer - `lost` lists distinct peptide sequences present in the original tryptic digest and absent from the edited digest. `gained` lists those present only in the edited digest. - Sort each list in ASCII order. Each sequence appears once; an empty difference is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "lost": [<amino-acid string>], "gained": [<amino-acid string>] } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Apply the DNA edits and report which complete tryptic peptide sequences are lost and gained relative to the original CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "edits": [ { "replace": [4, 6, "GCA"] } ] } ``` ## Rules - Use NCBI genetic code 11 with first-stop termination. - Apply the supplied edits simultaneously to the original DNA coordinates. - Translate and completely digest both versions under `trypsin_KR_notP_v1`, with zero missed cleavages. - Compare peptide sequence sets, ignoring copy number. ## Answer - `lost` lists distinct peptide sequences present in the original tryptic digest and absent from the edited digest. `gained` lists those present only in the edited digest. - Sort each list in ASCII order. Each sequence appears once; an empty difference is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "lost": [<amino-acid string>], "gained": [<amino-acid string>] } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "lost": ["AIFVLK", "MK"], "gained": ["MAAIFVLK"] }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Apply the DNA edits and report which complete tryptic peptide sequences are lost and gained relative to the original CDS. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "edits": {{edits}} } ``` ## Rules - Use NCBI genetic code 11 with first-stop termination. - Apply the supplied edits simultaneously to the original DNA coordinates. - Translate and completely digest both versions under `trypsin_KR_notP_v1`, with zero missed cleavages. - Compare peptide sequence sets, ignoring copy number. ## Answer - `lost` lists distinct peptide sequences present in the original tryptic digest and absent from the edited digest. `gained` lists those present only in the edited digest. - Sort each list in ASCII order. Each sequence appears once; an empty difference is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "lost": [<amino-acid string>], "gained": [<amino-acid string>] } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `lost` lists distinct peptide sequences present in the original tryptic digest and absent from the edited digest. `gained` lists those present only in the edited digest. Sort each list in ASCII order. Each sequence appears once; an empty difference is []. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "lost": [<amino-acid string>], "gained": [<amino-acid string>] } Steps and final output Translate and digest both the original and edited CDS. Report lost peptides as sequences present only in the original digest, and gained peptides as sequences present only in the edited digest. Count repeated copies only once for these two comparisons. Output: JSON object with exactly lost, gained Literal task fields { "composition_type": "pipeline.variant_peptides", "output_format": "json", "scorer": "exact_match", "mode": "branch", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Reference material. Initiation after DNA editing, NCBI genetic code 11: internal DNA codons, Trypsin cleavage specificity
From specified RNA edits to protein`pipeline.rna_edit_translation`	Transcribe a CDS, apply supplied RNA edits, and translate the edited RNA to obtain the final protein.
From specified RNA edits to protein The CDS gives the initial coding RNA, and the question states the exact RNA edits. Apply those edits before translating. The original DNA is retained as the source, but it must not be used in place of the edited RNA at the final step. This tests the consequence of a change introduced specifically at the RNA stage. Practical context ApoB provides a direct biological counterpart: APOBEC1-mediated C-to-U editing turns a CAA glutamine codon into a UAA stop, producing a shorter protein. The worked example uses that same codon change on a small illustrative CDS. The edit is supplied; predicting the enzyme’s site selection is outside the task. Method references: ApoB RNA editing: CAA becomes a stop codon, The apoB-editing catalytic subunit is a cytidine deaminase. Component tasks Make the coding RNA`dna.transcribe`Apply specified RNA edits`rna.apply_edits`Translate coding RNA`rna.translate` Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CAA ACT TCC TGA`; `rna_edits`: `[{"replace": [34, 34, "U"]}]`. Procedure. Transcribe the R12Q substrate, then change RNA C34 to U. CAA at codon 12 becomes UAA, so translation stops after MKAIFVLKGWW and excludes all downstream codons. Final answer: MKAIFVLKGWW Checking the answer. The final amino-acid sequence must be the translation of the edited RNA under the specified start and stop conventions. Easy example · references supplied Prompt ## Task Apply these specified edits to the coding RNA of this CDS, then return the edited RNA’s translated protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CAA ACT TCC TGA", "rna_edits": [ { "replace": [34, 34, "U"] } ] } ``` ## Rules - Transcribe the CDS and apply the supplied edits using the original RNA coordinates. - Translate the edited RNA from its first base with NCBI genetic code 11. Do not search for a new start position after an edit. - Stop at the first in-frame stop, omit its symbol, and ignore an incomplete trailing codon. - If the first edited codon is not an allowed initiator, use its ordinary internal codon meaning. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Initiation of coding RNA At the first RNA codon, UUG, CUG, AUU, AUC, AUA, AUG, and GUG encode initiating M under NCBI genetic code 11. If that first codon is not an allowed initiator, use its ordinary internal assignment. ### NCBI genetic code 11: internal RNA codons Each internal RNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text UUU=F UUC=F UUA=L UUG=L UCU=S UCC=S UCA=S UCG=S UAU=Y UAC=Y UAA=STOP UAG=STOP UGU=C UGC=C UGA=STOP UGG=W CUU=L CUC=L CUA=L CUG=L CCU=P CCC=P CCA=P CCG=P CAU=H CAC=H CAA=Q CAG=Q CGU=R CGC=R CGA=R CGG=R AUU=I AUC=I AUA=I AUG=M ACU=T ACC=T ACA=T ACG=T AAU=N AAC=N AAA=K AAG=K AGU=S AGC=S AGA=R AGG=R GUU=V GUC=V GUA=V GUG=V GCU=A GCC=A GCA=A GCG=A GAU=D GAC=D GAA=E GAG=E GGU=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: MKAIFVLKGWW` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL with R12Q (CGC→CAA), followed by a supplied C34→U RNA edit converting CAA to UAA. It illustrates a premature stop; no native trpL editing event is claimed. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Apply these specified edits to the coding RNA of this CDS, then return the edited RNA’s translated protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "rna_edits": {{rna_edits}} } ``` ## Rules - Transcribe the CDS and apply the supplied edits using the original RNA coordinates. - Translate the edited RNA from its first base with NCBI genetic code 11. Do not search for a new start position after an edit. - Stop at the first in-frame stop, omit its symbol, and ignore an incomplete trailing codon. - If the first edited codon is not an allowed initiator, use its ordinary internal codon meaning. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Initiation of coding RNA At the first RNA codon, UUG, CUG, AUU, AUC, AUA, AUG, and GUG encode initiating M under NCBI genetic code 11. If that first codon is not an allowed initiator, use its ordinary internal assignment. ### NCBI genetic code 11: internal RNA codons Each internal RNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text UUU=F UUC=F UUA=L UUG=L UCU=S UCC=S UCA=S UCG=S UAU=Y UAC=Y UAA=STOP UAG=STOP UGU=C UGC=C UGA=STOP UGG=W CUU=L CUC=L CUA=L CUG=L CCU=P CCC=P CCA=P CCG=P CAU=H CAC=H CAA=Q CAG=Q CGU=R CGC=R CGA=R CGG=R AUU=I AUC=I AUA=I AUG=M ACU=T ACC=T ACA=T ACG=T AAU=N AAC=N AAA=K AAG=K AGU=S AGC=S AGA=R AGG=R GUU=V GUC=V GUA=V GUG=V GCU=A GCC=A GCA=A GCG=A GAU=D GAC=D GAA=E GAG=E GGU=G GGC=G GGA=G GGG=G ``` ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Apply these specified edits to the coding RNA of this CDS, then return the edited RNA’s translated protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CAA ACT TCC TGA", "rna_edits": [ { "replace": [34, 34, "U"] } ] } ``` ## Rules - Transcribe the CDS and apply the supplied edits using the original RNA coordinates. - Translate the edited RNA from its first base with NCBI genetic code 11. Do not search for a new start position after an edit. - Stop at the first in-frame stop, omit its symbol, and ignore an incomplete trailing codon. - If the first edited codon is not an allowed initiator, use its ordinary internal codon meaning. ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: MKAIFVLKGWW` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL with R12Q (CGC→CAA), followed by a supplied C34→U RNA edit converting CAA to UAA. It illustrates a premature stop; no native trpL editing event is claimed. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Apply these specified edits to the coding RNA of this CDS, then return the edited RNA’s translated protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "rna_edits": {{rna_edits}} } ``` ## Rules - Transcribe the CDS and apply the supplied edits using the original RNA coordinates. - Translate the edited RNA from its first base with NCBI genetic code 11. Do not search for a new start position after an edit. - Stop at the first in-frame stop, omit its symbol, and ignore an incomplete trailing codon. - If the first edited codon is not an allowed initiator, use its ordinary internal codon meaning. ## Answer - Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. - An empty sequence is the token EMPTY_SEQUENCE. After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase one-letter amino-acid sequence, ending before the first in-frame stop. An empty sequence is the token EMPTY_SEQUENCE. Steps and final output Transcribe the CDS, apply the supplied RNA edits, then translate the edited RNA. Start at the original reading origin and stop at the first in-frame stop codon. Output: post-edit protein string Literal task fields { "composition_type": "pipeline.rna_edit_translation", "output_format": "scalar", "scorer": "exact_match", "mode": "pipeline", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Reference material. Coding DNA to RNA, Initiation of coding RNA, NCBI genetic code 11: internal RNA codons
From specified RNA edits to RNA fragments`pipeline.rna_edit_digest`	Apply specified edits to a CDS-derived RNA and determine the RNA fragments produced afterward by RNase T1.
From specified RNA edits to RNA fragments Transcribe the CDS, apply the supplied RNA edits, then use the sequence-only RNase T1 rule on that edited RNA. Editing can change a fragment’s letters even when the cuts stay fixed; literal substitutions that add or remove G can also change cut positions. This composition uses explicit A/C/G/U letters. Treating inosine as G for translation does not establish RNase T1 cleavage chemistry at inosine. Practical context A supplied RNA edit changes the sequence fingerprint expected from a real RNase T1 digest. The example uses C→U, as in apoB editing: it changes letters within a fragment without changing the G-dependent cuts. This is sequence interpretation for RNA characterization, with mass spectrometry providing a possible downstream readout. Method references: RNase T1: enzyme specificity and applications, RNA and mRNA sequence mapping by mass spectrometry. Component tasks Make the coding RNA`dna.transcribe`Apply specified RNA edits`rna.apply_edits`Digest coding RNA with RNase T1 rules`rna.t1_digest` Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CAA ACT TCC TGA`; `rna_edits`: `[{"replace": [34, 34, "U"]}]`. Procedure. Apply the same C34→U edit to the complete RNA, then cut after every G under the RNase T1 rule. Translation stopping at UAA does not truncate the RNA used for this digest. Final answer: [ { "start": 1, "end": 3, "fragment": "AUG" }, { "start": 4, "end": 7, "fragment": "AAAG" }, { "start": 8, "end": 16, "fragment": "CAAUUUUCG" }, { "start": 17, "end": 21, "fragment": "UACUG" }, { "start": 22, "end": 25, "fragment": "AAAG" }, { "start": 26, "end": 26, "fragment": "G" }, { "start": 27, "end": 29, "fragment": "UUG" }, { "start": 30, "end": 30, "fragment": "G" }, { "start": 31, "end": 32, "fragment": "UG" }, { "start": 33, "end": 33, "fragment": "G" }, { "start": 34, "end": 44, "fragment": "UAAACUUCCUG" }, { "start": 45, "end": 45, "fragment": "A" } ] Checking the answer. Fragment sequences and intervals must describe the edited RNA’s digest, not the original RNA’s digest. Easy example · references supplied Prompt ## Task Apply the RNA edits, then return the complete RNase T1 digest of the edited RNA. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CAA ACT TCC TGA", "rna_edits": [ { "replace": [34, 34, "U"] } ] } ``` ## Rules - Transcribe the CDS and apply the supplied edits in original RNA coordinates. - Digest the resulting RNA completely with RNase T1, with zero missed eligible cuts. - Treat every position as accessible. Do not use a pairing map to block any otherwise eligible cut. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNase T1 cleavage specificity RNase T1 cleaves RNA after G. That G remains on the upstream fragment. ## Answer - `start` and `end` are 1-based inclusive positions in the edited RNA; `fragment` is the continuous uppercase RNA sequence. - Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer Final answer: [ { "start": 1, "end": 3, "fragment": "AUG" }, { "start": 4, "end": 7, "fragment": "AAAG" }, { "start": 8, "end": 16, "fragment": "CAAUUUUCG" }, { "start": 17, "end": 21, "fragment": "UACUG" }, { "start": 22, "end": 25, "fragment": "AAAG" }, { "start": 26, "end": 26, "fragment": "G" }, { "start": 27, "end": 29, "fragment": "UUG" }, { "start": 30, "end": 30, "fragment": "G" }, { "start": 31, "end": 32, "fragment": "UG" }, { "start": 33, "end": 33, "fragment": "G" }, { "start": 34, "end": 44, "fragment": "UAAACUUCCUG" }, { "start": 45, "end": 45, "fragment": "A" } ] Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL with R12Q (CGC→CAA), followed by a supplied C34→U RNA edit converting CAA to UAA. It illustrates a premature stop; no native trpL editing event is claimed. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Apply the RNA edits, then return the complete RNase T1 digest of the edited RNA. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "rna_edits": {{rna_edits}} } ``` ## Rules - Transcribe the CDS and apply the supplied edits in original RNA coordinates. - Digest the resulting RNA completely with RNase T1, with zero missed eligible cuts. - Treat every position as accessible. Do not use a pairing map to block any otherwise eligible cut. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNase T1 cleavage specificity RNase T1 cleaves RNA after G. That G remains on the upstream fragment. ## Answer - `start` and `end` are 1-based inclusive positions in the edited RNA; `fragment` is the continuous uppercase RNA sequence. - Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Apply the RNA edits, then return the complete RNase T1 digest of the edited RNA. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CAA ACT TCC TGA", "rna_edits": [ { "replace": [34, 34, "U"] } ] } ``` ## Rules - Transcribe the CDS and apply the supplied edits in original RNA coordinates. - Digest the resulting RNA completely with RNase T1, with zero missed eligible cuts. - Treat every position as accessible. Do not use a pairing map to block any otherwise eligible cut. ## Answer - `start` and `end` are 1-based inclusive positions in the edited RNA; `fragment` is the continuous uppercase RNA sequence. - Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer Final answer: [ { "start": 1, "end": 3, "fragment": "AUG" }, { "start": 4, "end": 7, "fragment": "AAAG" }, { "start": 8, "end": 16, "fragment": "CAAUUUUCG" }, { "start": 17, "end": 21, "fragment": "UACUG" }, { "start": 22, "end": 25, "fragment": "AAAG" }, { "start": 26, "end": 26, "fragment": "G" }, { "start": 27, "end": 29, "fragment": "UUG" }, { "start": 30, "end": 30, "fragment": "G" }, { "start": 31, "end": 32, "fragment": "UG" }, { "start": 33, "end": 33, "fragment": "G" }, { "start": 34, "end": 44, "fragment": "UAAACUUCCUG" }, { "start": 45, "end": 45, "fragment": "A" } ] Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Native trpL with R12Q (CGC→CAA), followed by a supplied C34→U RNA edit converting CAA to UAA. It illustrates a premature stop; no native trpL editing event is claimed. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Apply the RNA edits, then return the complete RNase T1 digest of the edited RNA. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "rna_edits": {{rna_edits}} } ``` ## Rules - Transcribe the CDS and apply the supplied edits in original RNA coordinates. - Digest the resulting RNA completely with RNase T1, with zero missed eligible cuts. - Treat every position as accessible. Do not use a pairing map to block any otherwise eligible cut. ## Answer - `start` and `end` are 1-based inclusive positions in the edited RNA; `fragment` is the continuous uppercase RNA sequence. - Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `start` and `end` are 1-based inclusive positions in the edited RNA; `fragment` is the continuous uppercase RNA sequence. Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] Steps and final output Digest the edited RNA with the RNase T1 rule. This question treats every G as accessible (accessibility=all, pairs=[]); it does not reuse a pairing map from before the edits. Output: JSON array of records in the specified order Literal task fields { "composition_type": "pipeline.rna_edit_digest", "output_format": "json", "scorer": "exact_match", "mode": "pipeline", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Reference material. Coding DNA to RNA, RNase T1 cleavage specificity
From a pairing annotation to eligible cuts`pipeline.structure_gated_digest`	Use a supplied RNA-structure annotation to decide which RNase T1 cuts are allowed in coding RNA.
From a pairing annotation to eligible cuts The question supplies a CDS and a dot-bracket annotation of its coding RNA. Convert the annotation into paired positions, then permit cleavage only at G bases marked unpaired. The structure stays fixed for the calculation; the model is not asked to refold fragments after cutting. Exercise scope RNase T1 preferentially cleaves single-stranded G sites; RNA protection assays exploit the difference between exposed and protected RNA. That gives the composition a biochemical basis. Its fixed pairing map and binary cut/no-cut mask simplify actual accessibility and reaction kinetics rather than predicting an experimental digestion pattern. Method reference: RNase T1: enzyme specificity and applications. Component tasks Make the coding RNA`dna.transcribe`Decode a pairing map`rna.dotbracket_to_pairs`Digest coding RNA with RNase T1 rules`rna.t1_digest` Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `dot_bracket`: `...(((....)))...........((.......).).........`. Procedure. Decode the supplied annotation and mark every paired nucleotide. Cut after G only if that G is unpaired; G25 and G26 are protected, so their neighboring digest intervals merge. Final answer: [ { "start": 1, "end": 3, "fragment": "AUG" }, { "start": 4, "end": 7, "fragment": "AAAG" }, { "start": 8, "end": 16, "fragment": "CAAUUUUCG" }, { "start": 17, "end": 21, "fragment": "UACUG" }, { "start": 22, "end": 29, "fragment": "AAAGGUUG" }, { "start": 30, "end": 30, "fragment": "G" }, { "start": 31, "end": 32, "fragment": "UG" }, { "start": 33, "end": 33, "fragment": "G" }, { "start": 34, "end": 35, "fragment": "CG" }, { "start": 36, "end": 44, "fragment": "CACUUCCUG" }, { "start": 45, "end": 45, "fragment": "A" } ] Checking the answer. Every returned cut must satisfy both the sequence rule and the supplied structural eligibility rule. Easy example · references supplied Prompt ## Task Digest the coding RNA with RNase T1, allowing cuts only at positions that are unpaired in the supplied structure. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "dot_bracket": "...(((....)))...........((.......).)........." } ``` ## Rules - Transcribe the CDS and decode the supplied dot-bracket annotation into paired positions. - Apply the RNase T1 specificity rule only where the recognized nucleotide is unpaired in that map. Cut every eligible site. - Use the original supplied map throughout digestion. Do not recompute it after making a cut. - Keep every nonempty fragment in its original RNA order. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Dot-bracket notation Matching parentheses mark paired RNA positions; a dot marks an unpaired position. Each closing parenthesis pairs with the most recent unmatched opening parenthesis. ### RNase T1 cleavage specificity RNase T1 cleaves RNA after G. That G remains on the upstream fragment. ## Answer - `start` and `end` are 1-based inclusive positions in the original CDS-derived RNA; `fragment` is the continuous uppercase RNA sequence. - Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer Final answer: [ { "start": 1, "end": 3, "fragment": "AUG" }, { "start": 4, "end": 7, "fragment": "AAAG" }, { "start": 8, "end": 16, "fragment": "CAAUUUUCG" }, { "start": 17, "end": 21, "fragment": "UACUG" }, { "start": 22, "end": 29, "fragment": "AAAGGUUG" }, { "start": 30, "end": 30, "fragment": "G" }, { "start": 31, "end": 32, "fragment": "UG" }, { "start": 33, "end": 33, "fragment": "G" }, { "start": 34, "end": 35, "fragment": "CG" }, { "start": 36, "end": 44, "fragment": "CACUUCCUG" }, { "start": 45, "end": 45, "fragment": "A" } ] Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The fixed pairing annotation is supplied for the exercise, not inferred from trpL biology. It protects the paired G bases at positions 25 and 26 from the stated RNase T1 rule. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Digest the coding RNA with RNase T1, allowing cuts only at positions that are unpaired in the supplied structure. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "dot_bracket": "{{dot_bracket}}" } ``` ## Rules - Transcribe the CDS and decode the supplied dot-bracket annotation into paired positions. - Apply the RNase T1 specificity rule only where the recognized nucleotide is unpaired in that map. Cut every eligible site. - Use the original supplied map throughout digestion. Do not recompute it after making a cut. - Keep every nonempty fragment in its original RNA order. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Dot-bracket notation Matching parentheses mark paired RNA positions; a dot marks an unpaired position. Each closing parenthesis pairs with the most recent unmatched opening parenthesis. ### RNase T1 cleavage specificity RNase T1 cleaves RNA after G. That G remains on the upstream fragment. ## Answer - `start` and `end` are 1-based inclusive positions in the original CDS-derived RNA; `fragment` is the continuous uppercase RNA sequence. - Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Digest the coding RNA with RNase T1, allowing cuts only at positions that are unpaired in the supplied structure. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "dot_bracket": "...(((....)))...........((.......).)........." } ``` ## Rules - Transcribe the CDS and decode the supplied dot-bracket annotation into paired positions. - Apply the RNase T1 specificity rule only where the recognized nucleotide is unpaired in that map. Cut every eligible site. - Use the original supplied map throughout digestion. Do not recompute it after making a cut. - Keep every nonempty fragment in its original RNA order. ## Answer - `start` and `end` are 1-based inclusive positions in the original CDS-derived RNA; `fragment` is the continuous uppercase RNA sequence. - Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer Final answer: [ { "start": 1, "end": 3, "fragment": "AUG" }, { "start": 4, "end": 7, "fragment": "AAAG" }, { "start": 8, "end": 16, "fragment": "CAAUUUUCG" }, { "start": 17, "end": 21, "fragment": "UACUG" }, { "start": 22, "end": 29, "fragment": "AAAGGUUG" }, { "start": 30, "end": 30, "fragment": "G" }, { "start": 31, "end": 32, "fragment": "UG" }, { "start": 33, "end": 33, "fragment": "G" }, { "start": 34, "end": 35, "fragment": "CG" }, { "start": 36, "end": 44, "fragment": "CACUUCCUG" }, { "start": 45, "end": 45, "fragment": "A" } ] Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The fixed pairing annotation is supplied for the exercise, not inferred from trpL biology. It protects the paired G bases at positions 25 and 26 from the stated RNase T1 rule. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Digest the coding RNA with RNase T1, allowing cuts only at positions that are unpaired in the supplied structure. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "dot_bracket": "{{dot_bracket}}" } ``` ## Rules - Transcribe the CDS and decode the supplied dot-bracket annotation into paired positions. - Apply the RNase T1 specificity rule only where the recognized nucleotide is unpaired in that map. Cut every eligible site. - Use the original supplied map throughout digestion. Do not recompute it after making a cut. - Keep every nonempty fragment in its original RNA order. ## Answer - `start` and `end` are 1-based inclusive positions in the original CDS-derived RNA; `fragment` is the continuous uppercase RNA sequence. - Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `start` and `end` are 1-based inclusive positions in the original CDS-derived RNA; `fragment` is the continuous uppercase RNA sequence. Return every nonempty fragment in 5′→3′ order, retaining repeats and accounting for every base exactly once. An empty RNA gives []. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. [ { "start": <positive integer>, "end": <positive integer>, "fragment": <RNA string> } ] Steps and final output Transcribe the CDS and convert the supplied dot-bracket notation into paired positions. Use both to digest the RNA with accessibility=unpaired_only. Keep the supplied pairing map fixed during cleavage. Output: JSON array of records in the specified order Literal task fields { "composition_type": "pipeline.structure_gated_digest", "output_format": "json", "scorer": "exact_match", "mode": "pipeline", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Reference material. Coding DNA to RNA, Dot-bracket notation, RNase T1 cleavage specificity
Remove cloning sites while controlling GC`joint.clone_ready_cds`	Find one synonymous CDS that removes unwanted DNA sites and satisfies every supplied local GC requirement.
Remove cloning sites while controlling GC The input supplies forbidden patterns, a strand-search rule, and local GC limits. Both requirements apply to the same final CDS. Removing a site and then fixing GC can recreate the site, so a sequence is accepted only after all checks pass together. The encoded protein and fixed terminal codons must also be preserved. Practical context This combines two ordinary construct-design requirements: remove internal sites that would interfere with an assembly method and avoid unsuitable local GC while keeping the protein fixed. Practical constraint-based tools handle this kind of simultaneous repair, because fixing one feature can create another violation. Method references: GoldenMutagenesis domestication, DNA Chisel examples. Component tasks Remove specified DNA motifs`design.remove_dna_motifs`Control local GC by recoding`design.gc_windows`Compare encoded proteins`cds.protein_equivalence` Task Outline Inputs. `cds`: `ATG AAA GAA TTC TTC GTA CTG TGA`; `forbidden_patterns`: `["GAATTC"]`; `strand_mode`: `both`; `window`: `9`; `gc_bounds`: `["2/9", "2/3"]`. Procedure. For each synonymous cloning-substrate encoding, jointly require no GAATTC on either strand and compliant GC fractions in every nine-base window. Choose the first complete DNA satisfying both conditions. Final answer: ATGAAAGAGTTCTTCGTACTATGA Checking the answer. Any one candidate satisfying all final requirements is accepted. If none exists, the instance needs a certified IMPOSSIBLE answer. Easy example · references supplied Prompt ## Task Find one synonymous CDS that removes every forbidden DNA pattern and satisfies every GC-window bound. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. GC bounds are inclusive fractions. A string such as `"1/3"` denotes that exact rational value. ```json { "cds": "ATG AAA GAA TTC TTC GTA CTG TGA", "forbidden_patterns": ["GAATTC"], "strand_mode": "both", "window": 9, "gc_bounds": ["2/9", "2/3"] } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Remove every forbidden IUPAC pattern hit under `strand_mode`, including overlapping hits and hits spanning codons. - Move a window of length `window` over the full final CDS one base at a time. Every full window must satisfy the inclusive GC-fraction bounds. - Check the motif and GC conditions together on the same final sequence. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ### GC fraction GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGAGTTCTTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The minigene has A3E and I4F substitutions (GCA→GAA, ATT→TTC), creating the real EcoRI recognition sequence GAATTC at bases 7–12. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find one synonymous CDS that removes every forbidden DNA pattern and satisfies every GC-window bound. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. GC bounds are inclusive fractions. A string such as `"1/3"` denotes that exact rational value. ```json { "cds": "{{cds}}", "forbidden_patterns": {{forbidden_patterns}}, "strand_mode": "{{strand_mode}}", "window": {{window}}, "gc_bounds": {{gc_bounds}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Remove every forbidden IUPAC pattern hit under `strand_mode`, including overlapping hits and hits spanning codons. - Move a window of length `window` over the full final CDS one base at a time. Every full window must satisfy the inclusive GC-fraction bounds. - Check the motif and GC conditions together on the same final sequence. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ### GC fraction GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Find one synonymous CDS that removes every forbidden DNA pattern and satisfies every GC-window bound. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. GC bounds are inclusive fractions. A string such as `"1/3"` denotes that exact rational value. ```json { "cds": "ATG AAA GAA TTC TTC GTA CTG TGA", "forbidden_patterns": ["GAATTC"], "strand_mode": "both", "window": 9, "gc_bounds": ["2/9", "2/3"] } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Remove every forbidden IUPAC pattern hit under `strand_mode`, including overlapping hits and hits spanning codons. - Move a window of length `window` over the full final CDS one base at a time. Every full window must satisfy the inclusive GC-fraction bounds. - Check the motif and GC conditions together on the same final sequence. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGAGTTCTTCGTACTATGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The minigene has A3E and I4F substitutions (GCA→GAA, ATT→TTC), creating the real EcoRI recognition sequence GAATTC at bases 7–12. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find one synonymous CDS that removes every forbidden DNA pattern and satisfies every GC-window bound. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. GC bounds are inclusive fractions. A string such as `"1/3"` denotes that exact rational value. ```json { "cds": "{{cds}}", "forbidden_patterns": {{forbidden_patterns}}, "strand_mode": "{{strand_mode}}", "window": {{window}}, "gc_bounds": {{gc_bounds}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Remove every forbidden IUPAC pattern hit under `strand_mode`, including overlapping hits and hits spanning codons. - Move a window of length `window` over the full final CDS one base at a time. Every full window must satisfy the inclusive GC-fraction bounds. - Check the motif and GC conditions together on the same final sequence. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Steps and final output Find one CDS that removes the forbidden motifs and meets every GC-window bound. Check both conditions on the final sequence, verify that it encodes the original protein, and keep the first and last codons unchanged. A later rewrite must not undo an earlier requirement. Output: canonical DNA string, or IMPOSSIBLE Literal task fields { "composition_type": "joint.clone_ready_cds", "output_format": "scalar", "scorer": "exact_match", "mode": "joint", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." } Reference material. Synonymous codons, NCBI genetic code 11: internal DNA codons, IUPAC DNA base sets, DNA reverse complement, GC fraction
Repair RNA pairs and verify the protein`pipeline.repair_and_verify`	Repair specified coding-RNA pairs through synonymous DNA edits, then verify both pairing and protein preservation.
Repair RNA pairs and verify the protein The question gives the original CDS, intended RNA pairs, and positions allowed to change. Propose a synonymous repair, transcribe that candidate, and check its pairs. Independently compare the candidate’s protein with the original protein. This makes both conditions explicit rather than treating one as a consequence of the other. Practical context The rpsT compensatory-mutation study supplies a concrete precedent for repairing proposed coding-RNA contacts through synonymous changes. This composition verifies that the designed sequence restores the specified pairs and retains the protein. A resulting sequence is a candidate experimental control; the pair check alone cannot establish its fitness effect. Method reference: Synonymous mutations and restoration of mRNA base pairing. Component tasks Repair pairing synonymously`design.pairing_repair`Make the coding RNA`dna.transcribe`Check specified RNA pairs`rna.pair_compatibility`Compare encoded proteins`cds.protein_equivalence` Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `pairs`: `[[3, 9]]`; `editable_positions`: `[9]`. Procedure. Repair the supplied G3–A9 contact using only DNA base 9, then translate that same final CDS. GCC supplies the required C9 while preserving the original MKAIFVL protein. Final answer: ATGAAAGCCATTTTCGTACTGTGA Checking the answer. One final CDS must pass the RNA-pair check, protein-equivalence check, and edit permissions. Easy example · references supplied Prompt ## Task Repair the listed RNA pairs through permitted DNA substitutions, and verify that the final CDS still encodes the original protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "pairs": [ [3, 9] ], "editable_positions": [9] } ``` ## Rules - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Every listed pair must be canonical in the final coding RNA, with no wobble pairing. Do not infer unlisted pairs. - Verify that this same final DNA still encodes the complete original protein under NCBI genetic code 11. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCCATTTTCGTACTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The supplied pair (3,9) is initially G–A; the only permitted DNA position is the third base of Ala codon 3. The map is an illustrative repair target. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Repair the listed RNA pairs through permitted DNA substitutions, and verify that the final CDS still encodes the original protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pairs": {{pairs}}, "editable_positions": {{editable_positions}} } ``` ## Rules - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Every listed pair must be canonical in the final coding RNA, with no wobble pairing. Do not infer unlisted pairs. - Verify that this same final DNA still encodes the complete original protein under NCBI genetic code 11. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Repair the listed RNA pairs through permitted DNA substitutions, and verify that the final CDS still encodes the original protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "pairs": [ [3, 9] ], "editable_positions": [9] } ``` ## Rules - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Every listed pair must be canonical in the final coding RNA, with no wobble pairing. Do not infer unlisted pairs. - Verify that this same final DNA still encodes the complete original protein under NCBI genetic code 11. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCCATTTTCGTACTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The supplied pair (3,9) is initially G–A; the only permitted DNA position is the third base of Ala codon 3. The map is an illustrative repair target. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Repair the listed RNA pairs through permitted DNA substitutions, and verify that the final CDS still encodes the original protein. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "pairs": {{pairs}}, "editable_positions": {{editable_positions}} } ``` ## Rules - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Only DNA positions in `editable_positions` may change. - Every listed pair must be canonical in the final coding RNA, with no wobble pairing. Do not infer unlisted pairs. - Verify that this same final DNA still encodes the complete original protein under NCBI genetic code 11. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. If no valid result exists, return: IMPOSSIBLE Steps and final output Transcribe the repaired CDS and check every supplied RNA pair. Translate the repaired and original CDSs to verify that they encode the same protein. Both checks must pass for the same final DNA sequence. Output: canonical DNA string, or IMPOSSIBLE Literal task fields { "composition_type": "pipeline.repair_and_verify", "output_format": "scalar", "scorer": "exact_match", "mode": "branch", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." } Reference material. Synonymous codons, NCBI genetic code 11: internal DNA codons, Coding DNA to RNA, Canonical RNA base pairs
Balance an RNA structural constraint and codon preference`joint.structure_and_codons`	Within a changed-base budget, preserve the protein and meet an RNA pair-score threshold while maximizing codon preference.
Balance an RNA structural constraint and codon preference The input gives allowed DNA-change counts, a fixed RNA pair map, pair scores, a minimum total pair score, and codon-preference weights. First define feasible synonymous candidates: each must satisfy the edit range and structural threshold. Among those candidates, maximize codon adaptation index. The structure score is a pass/fail constraint here, not a second objective to maximize. Exercise scope The composition models the structure of a multi-objective design problem, but its supplied pair-score threshold is artificial. LinearDesign is a practical counterpart that combines codon adaptation with a thermodynamic RNA-folding objective. The benchmark tests constraint coordination without claiming to reproduce that stability model. Method reference: LinearDesign. Component tasks Rewrite with a specified number of DNA changes`design.synonymous_distance`Make the coding RNA`dna.transcribe`Score a specified RNA structure`rna.structure_score`Calculate codon adaptation index`stats.cai` Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `edit_bounds`: `[2, 2]`; `pairs`: `[[9, 18]]`; `pair_scores`: `{"AU": 2, "UA": 2, "CG": 3, "GC": 3}`; `min_pair_score`: `2`; `weights`: supplied table shown in the prompt. Procedure. Keep candidates with exactly two changed bases and a canonical pair between RNA positions 9 and 18 scoring at least two. Maximize the product of internal codon weights among those candidates, then apply the DNA tie-break. Final answer: ATGAAAGCCATTTTCGTGCTGTGA Checking the answer. The accepted candidate must satisfy every feasibility condition and attain the best codon score among feasible candidates. Easy example · references supplied Prompt ## Task Find a synonymous CDS within the edit budget that meets the RNA pair-score threshold and has the highest permitted CAI. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "edit_bounds": [2, 2], "pairs": [ [9, 18] ], "pair_scores": { "AU": 2, "UA": 2, "CG": 3, "GC": 3 }, "min_pair_score": 2, "weights": { "AAA": 1.0, "AAG": 0.3, "GCA": 0.2, "GCC": 0.3, "GCG": 1.0, "GCT": 0.5, "ATA": 0.2, "ATC": 0.3, "ATT": 1.0, "TTC": 1.0, "TTT": 0.3, "GTA": 0.2, "GTC": 0.3, "GTG": 1.0, "GTT": 0.5, "CTA": 0.2, "CTC": 0.3, "CTG": 1.0, "CTT": 0.5, "TTA": 0.6, "TTG": 0.7 } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Count changed nucleotide positions relative to the original CDS. The count must meet both inclusive `edit_bounds`. - Keep the supplied RNA pair map fixed. Every listed pair must be canonical, and the sum of its supplied pair scores must be at least `min_pair_score`. - Among candidates meeting all constraints, maximize CAI using `weights`, excluding the initiator and terminal stop. ## Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ### Codon adaptation index For n included codons with supplied positive weights w₁,…,wₙ, CAI = (w₁ × … × wₙ)^(1/n), the geometric mean of those weights. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCCATTTTCGTGCTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Pair map and codon weights are supplied illustrative inputs. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find a synonymous CDS within the edit budget that meets the RNA pair-score threshold and has the highest permitted CAI. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "edit_bounds": {{edit_bounds}}, "pairs": {{pairs}}, "pair_scores": {{pair_scores}}, "min_pair_score": {{min_pair_score}}, "weights": {{weights}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Count changed nucleotide positions relative to the original CDS. The count must meet both inclusive `edit_bounds`. - Keep the supplied RNA pair map fixed. Every listed pair must be canonical, and the sum of its supplied pair scores must be at least `min_pair_score`. - Among candidates meeting all constraints, maximize CAI using `weights`, excluding the initiator and terminal stop. ## Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Canonical RNA base pairs The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C. ### Codon adaptation index For n included codons with supplied positive weights w₁,…,wₙ, CAI = (w₁ × … × wₙ)^(1/n), the geometric mean of those weights. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Find a synonymous CDS within the edit budget that meets the RNA pair-score threshold and has the highest permitted CAI. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "edit_bounds": [2, 2], "pairs": [ [9, 18] ], "pair_scores": { "AU": 2, "UA": 2, "CG": 3, "GC": 3 }, "min_pair_score": 2, "weights": { "AAA": 1.0, "AAG": 0.3, "GCA": 0.2, "GCC": 0.3, "GCG": 1.0, "GCT": 0.5, "ATA": 0.2, "ATC": 0.3, "ATT": 1.0, "TTC": 1.0, "TTT": 0.3, "GTA": 0.2, "GTC": 0.3, "GTG": 1.0, "GTT": 0.5, "CTA": 0.2, "CTC": 0.3, "CTG": 1.0, "CTT": 0.5, "TTA": 0.6, "TTG": 0.7 } } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Count changed nucleotide positions relative to the original CDS. The count must meet both inclusive `edit_bounds`. - Keep the supplied RNA pair map fixed. Every listed pair must be canonical, and the sum of its supplied pair scores must be at least `min_pair_score`. - Among candidates meeting all constraints, maximize CAI using `weights`, excluding the initiator and terminal stop. ## Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGCCATTTTCGTGCTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Pair map and codon weights are supplied illustrative inputs. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Find a synonymous CDS within the edit budget that meets the RNA pair-score threshold and has the highest permitted CAI. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "edit_bounds": {{edit_bounds}}, "pairs": {{pairs}}, "pair_scores": {{pair_scores}}, "min_pair_score": {{min_pair_score}}, "weights": {{weights}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - Count changed nucleotide positions relative to the original CDS. The count must meet both inclusive `edit_bounds`. - Keep the supplied RNA pair map fixed. Every listed pair must be canonical, and the sum of its supplied pair scores must be at least `min_pair_score`. - Among candidates meeting all constraints, maximize CAI using `weights`, excluding the initiator and terminal stop. ## Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding. If no valid result exists, return: IMPOSSIBLE Steps and final output Keep the protein unchanged and obey edit_bounds. The candidate’s coding RNA must satisfy every supplied pair and reach min_pair_score under wc_additive_v1. Among candidates meeting those conditions, maximize CAI using the supplied codon weights. Check all conditions on that same sequence. Output: canonical DNA string, or IMPOSSIBLE Literal task fields { "composition_type": "joint.structure_and_codons", "output_format": "scalar", "scorer": "exact_match", "mode": "joint", "selection_rule": "After achieving the stated optimum, break ties by the lexicographically smallest full DNA string, with A<C<G<T. Compare objective values before rounding." } Reference material. Synonymous codons, NCBI genetic code 11: internal DNA codons, Coding DNA to RNA, Canonical RNA base pairs, Codon adaptation index
Choose an RNA probe and a matched escape control`pipeline.target_and_escape`	Choose a target-specific RNA probe, then make a synonymous version of the target that no longer closely matches that same probe.
Choose an RNA probe and a matched escape control Use the target and comparison CDSs to select the required probe by the probe-selection task’s rules. Keep that selected probe fixed while redesigning the target CDS as an escape control. The redesigned target must preserve its protein and meet the supplied mismatch minimum at every RNA window. The comparisons stay unchanged. Practical context This provides a sequence-level analogue of an RNAi target and rescue control: the original RNA matches the reagent, while a synonymous construct preserves the protein and disrupts that match. Actual knockdown and rescue depend on reagent activity; a mismatch threshold is only the design screen. Method reference: RNAi-resistant rescue in a B-cyclin study. Component tasks Make the coding RNA`dna.transcribe`Select a CDS-specific RNA probe`assay.select_rna_probe`Match an RNA guide`rna.guide_hits`Create a guide-escape control`design.guide_escape` Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `background_cds_panel`: `["ATG AAA GCA GTT TTC GTA CTG TGA"]`; `probe_length`: `12`; `background_mismatch_floor`: `1`; `escape_mismatches`: `1`; `editable_positions`: `[6, 9, 12, 15, 18, 21]`. Procedure. First select the same target-specific 12-base probe as the probe-selection task. Keep it fixed, rewrite only permitted synonymous positions, and require every target window to have at least one mismatch to that probe’s reverse complement. Final answer: { "status": "ok", "probe": "AAAAUUGCUUUC", "escape_cds": "ATGAAAGCAATATTCGTACTATGA" } Checking the answer. Check probe selection and escape design separately, then verify that the reported probe is the one used to construct the control. Easy example · references supplied Prompt ## Task Choose a target-specific RNA probe, then design a synonymous version of the target that escapes that same probe. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "background_cds_panel": ["ATG AAA GCA GTT TTC GTA CTG TGA"], "probe_length": 12, "background_mismatch_floor": 1, "escape_mismatches": 1, "editable_positions": [6, 9, 12, 15, 18, 21] } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - First transcribe the original target and comparison CDSs and select a probe of exactly `probe_length` bases. - The probe must be a perfect reverse complement of a target window. Every equal-length comparison-RNA window must differ from the probe’s reverse complement at at least `background_mismatch_floor` positions. - Keep that selected probe fixed while designing the escape CDS. Check every equal-length window of the candidate coding RNA against the same probe’s reverse complement; each must differ at at least `escape_mismatches` positions. - Only DNA positions in `editable_positions` may change. Preserve the original protein, DNA length, first codon, and terminal stop. Keep comparison sequences unchanged. - Use ungapped comparisons and no wobble. If the selected probe has no valid escape design, return the specified failure object; do not switch to a different probe. ## Selection First select the ASCII-lexicographically smallest probe meeting the original target/background requirements (A<C<G<U). Keep it fixed, then choose the lexicographically smallest feasible escape DNA (A<C<G<T). If that selected probe has no feasible escape, return impossible; do not try a later probe. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNA reverse complement RNA complementary bases are A↔U and C↔G. Complement each base, then reverse the sequence to write its complementary partner 5′→3′. ## Answer - `probe` is the selected continuous uppercase 5′→3′ RNA sequence. `escape_cds` is the final continuous uppercase synonymous DNA sequence. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "probe": <RNA string>, "escape_cds": <DNA string> } ``` If no valid result exists, return: ```json { "status": "impossible", "probe": null, "escape_cds": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "status": "ok", "probe": "AAAAUUGCUUUC", "escape_cds": "ATGAAAGCAATATTCGTACTATGA" }` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Comparison is a designed I4V minigene variant (ATT→GTT). Probes and escape changes are computed design exercises, not validated reagents. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Choose a target-specific RNA probe, then design a synonymous version of the target that escapes that same probe. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "background_cds_panel": {{background_cds_panel}}, "probe_length": {{probe_length}}, "background_mismatch_floor": {{background_mismatch_floor}}, "escape_mismatches": {{escape_mismatches}}, "editable_positions": {{editable_positions}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - First transcribe the original target and comparison CDSs and select a probe of exactly `probe_length` bases. - The probe must be a perfect reverse complement of a target window. Every equal-length comparison-RNA window must differ from the probe’s reverse complement at at least `background_mismatch_floor` positions. - Keep that selected probe fixed while designing the escape CDS. Check every equal-length window of the candidate coding RNA against the same probe’s reverse complement; each must differ at at least `escape_mismatches` positions. - Only DNA positions in `editable_positions` may change. Preserve the original protein, DNA length, first codon, and terminal stop. Keep comparison sequences unchanged. - Use ungapped comparisons and no wobble. If the selected probe has no valid escape design, return the specified failure object; do not switch to a different probe. ## Selection First select the ASCII-lexicographically smallest probe meeting the original target/background requirements (A<C<G<U). Keep it fixed, then choose the lexicographically smallest feasible escape DNA (A<C<G<T). If that selected probe has no feasible escape, return impossible; do not try a later probe. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNA reverse complement RNA complementary bases are A↔U and C↔G. Complement each base, then reverse the sequence to write its complementary partner 5′→3′. ## Answer - `probe` is the selected continuous uppercase 5′→3′ RNA sequence. `escape_cds` is the final continuous uppercase synonymous DNA sequence. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "probe": <RNA string>, "escape_cds": <DNA string> } ``` If no valid result exists, return: ```json { "status": "impossible", "probe": null, "escape_cds": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Choose a target-specific RNA probe, then design a synonymous version of the target that escapes that same probe. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "background_cds_panel": ["ATG AAA GCA GTT TTC GTA CTG TGA"], "probe_length": 12, "background_mismatch_floor": 1, "escape_mismatches": 1, "editable_positions": [6, 9, 12, 15, 18, 21] } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - First transcribe the original target and comparison CDSs and select a probe of exactly `probe_length` bases. - The probe must be a perfect reverse complement of a target window. Every equal-length comparison-RNA window must differ from the probe’s reverse complement at at least `background_mismatch_floor` positions. - Keep that selected probe fixed while designing the escape CDS. Check every equal-length window of the candidate coding RNA against the same probe’s reverse complement; each must differ at at least `escape_mismatches` positions. - Only DNA positions in `editable_positions` may change. Preserve the original protein, DNA length, first codon, and terminal stop. Keep comparison sequences unchanged. - Use ungapped comparisons and no wobble. If the selected probe has no valid escape design, return the specified failure object; do not switch to a different probe. ## Selection First select the ASCII-lexicographically smallest probe meeting the original target/background requirements (A<C<G<U). Keep it fixed, then choose the lexicographically smallest feasible escape DNA (A<C<G<T). If that selected probe has no feasible escape, return impossible; do not try a later probe. ## Answer - `probe` is the selected continuous uppercase 5′→3′ RNA sequence. `escape_cds` is the final continuous uppercase synonymous DNA sequence. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "probe": <RNA string>, "escape_cds": <DNA string> } ``` If no valid result exists, return: ```json { "status": "impossible", "probe": null, "escape_cds": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "status": "ok", "probe": "AAAAUUGCUUUC", "escape_cds": "ATGAAAGCAATATTCGTACTATGA" }` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Comparison is a designed I4V minigene variant (ATT→GTT). Probes and escape changes are computed design exercises, not validated reagents. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Choose a target-specific RNA probe, then design a synonymous version of the target that escapes that same probe. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "background_cds_panel": {{background_cds_panel}}, "probe_length": {{probe_length}}, "background_mismatch_floor": {{background_mismatch_floor}}, "escape_mismatches": {{escape_mismatches}}, "editable_positions": {{editable_positions}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - First transcribe the original target and comparison CDSs and select a probe of exactly `probe_length` bases. - The probe must be a perfect reverse complement of a target window. Every equal-length comparison-RNA window must differ from the probe’s reverse complement at at least `background_mismatch_floor` positions. - Keep that selected probe fixed while designing the escape CDS. Check every equal-length window of the candidate coding RNA against the same probe’s reverse complement; each must differ at at least `escape_mismatches` positions. - Only DNA positions in `editable_positions` may change. Preserve the original protein, DNA length, first codon, and terminal stop. Keep comparison sequences unchanged. - Use ungapped comparisons and no wobble. If the selected probe has no valid escape design, return the specified failure object; do not switch to a different probe. ## Selection First select the ASCII-lexicographically smallest probe meeting the original target/background requirements (A<C<G<U). Keep it fixed, then choose the lexicographically smallest feasible escape DNA (A<C<G<T). If that selected probe has no feasible escape, return impossible; do not try a later probe. ## Answer - `probe` is the selected continuous uppercase 5′→3′ RNA sequence. `escape_cds` is the final continuous uppercase synonymous DNA sequence. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "probe": <RNA string>, "escape_cds": <DNA string> } ``` If no valid result exists, return: ```json { "status": "impossible", "probe": null, "escape_cds": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `probe` is the selected continuous uppercase 5′→3′ RNA sequence. `escape_cds` is the final continuous uppercase synonymous DNA sequence. Selection First select the ASCII-lexicographically smallest probe meeting the original target/background requirements (A<C<G<U). Keep it fixed, then choose the lexicographically smallest feasible escape DNA (A<C<G<T). If that selected probe has no feasible escape, return impossible; do not try a later probe. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "status": "ok", "probe": <RNA string>, "escape_cds": <DNA string> } If no valid result exists, return: { "status": "impossible", "probe": null, "escape_cds": null } Steps and final output Transcribe the target and background CDSs, then select a 5′→3′ RNA probe. Use that same probe sequence for the guide-matching and escape-redesign tasks. Every window in the redesigned target must have at least escape_mismatches mismatches to the probe. Keep the background sequences unchanged. Output: JSON object with exactly status, probe, escape_cds Literal task fields { "composition_type": "pipeline.target_and_escape", "output_format": "json", "scorer": "exact_match", "mode": "pipeline", "selection_rule": "First select the ASCII-lexicographically smallest probe meeting the original target/background requirements (A<C<G<U). Keep it fixed, then choose the lexicographically smallest feasible escape DNA (A<C<G<T). If that selected probe has no feasible escape, return impossible; do not try a later probe." } Reference material. Synonymous codons, NCBI genetic code 11: internal DNA codons, Coding DNA to RNA, RNA reverse complement
Encode a protein change and remove cloning sites`joint.protein_edit_and_cloning`	Introduce one requested amino-acid change and remove forbidden DNA sites with the fewest total base substitutions.
Encode a protein change and remove cloning sites The requested mutation defines the target protein: one residue changes and all others remain original. Synonymous edits elsewhere are allowed to remove unwanted DNA sites. Search those choices together and minimize the total number of changed DNA bases, while preserving the fixed first and terminal stop codons. This differs from freezing the locally cheapest codon change before attempting cleanup. Practical context This turns a requested amino-acid change into a construct compatible with a chosen assembly method. The final translation checks the desired mutation, while motif checks ensure that unwanted restriction sites were not retained or created by the edits. Method references: GoldenMutagenesis domestication, Site-directed mutagenesis protocol. Component tasks Encode a requested amino-acid substitution`construct.minimal_aa_edit`Remove specified DNA motifs`design.remove_dna_motifs`Translate DNA to protein`cds.translate` Task Outline Inputs. `cds`: `ATG AAA GAT TTC TTC GTA CTG TGA`; `residue_index`: `3`; `target_aa`: `E`; `forbidden_patterns`: `["GAATTC"]`; `strand_mode`: `both`. Procedure. Replace Asp3 with Glu in the cloning substrate. GAA would create GAATTC with the next Phe codon; GAG makes the same requested Glu substitution without that site and with the minimum number of base changes. Final answer: ATGAAAGAGTTCTTCGTACTGTGA Checking the answer. The final CDS must encode exactly the requested mutant protein, contain no forbidden sites, and minimize total changes with the stated tie-break. Easy example · references supplied Prompt ## Task Encode the requested amino-acid substitution and remove the forbidden DNA patterns using the fewest total nucleotide substitutions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GAT TTC TTC GTA CTG TGA", "residue_index": 3, "target_aa": "E", "forbidden_patterns": ["GAATTC"], "strand_mode": "both" } ``` ## Rules - Change only the requested amino acid in the protein. Keep every other residue unchanged under NCBI genetic code 11. - Synonymous DNA substitutions at other codons are allowed when needed to remove forbidden patterns. Keep DNA length, the original first codon, and terminal stop unchanged. - Remove every overlapping or cross-codon forbidden IUPAC match under `strand_mode`. - Count changed nucleotides across the entire final CDS, including synonymous changes outside the codon for the requested substitution. ## Selection Minimize the total number of changed nucleotides permitted by this task; among equally minimal candidates, return the lexicographically smallest full DNA string (A<C<G<T). ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGAGTTCTTCGTACTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The A3E/I4F cloning substrate instead starts with Asp at position 3 (GAT). Its requested D3E edit must avoid introducing GAATTC. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Encode the requested amino-acid substitution and remove the forbidden DNA patterns using the fewest total nucleotide substitutions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "residue_index": {{residue_index}}, "target_aa": "{{target_aa}}", "forbidden_patterns": {{forbidden_patterns}}, "strand_mode": "{{strand_mode}}" } ``` ## Rules - Change only the requested amino acid in the protein. Keep every other residue unchanged under NCBI genetic code 11. - Synonymous DNA substitutions at other codons are allowed when needed to remove forbidden patterns. Keep DNA length, the original first codon, and terminal stop unchanged. - Remove every overlapping or cross-codon forbidden IUPAC match under `strand_mode`. - Count changed nucleotides across the entire final CDS, including synonymous changes outside the codon for the requested substitution. ## Selection Minimize the total number of changed nucleotides permitted by this task; among equally minimal candidates, return the lexicographically smallest full DNA string (A<C<G<T). ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Hard example · reference knowledge assumed Prompt ## Task Encode the requested amino-acid substitution and remove the forbidden DNA patterns using the fewest total nucleotide substitutions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GAT TTC TTC GTA CTG TGA", "residue_index": 3, "target_aa": "E", "forbidden_patterns": ["GAATTC"], "strand_mode": "both" } ``` ## Rules - Change only the requested amino acid in the protein. Keep every other residue unchanged under NCBI genetic code 11. - Synonymous DNA substitutions at other codons are allowed when needed to remove forbidden patterns. Keep DNA length, the original first codon, and terminal stop unchanged. - Remove every overlapping or cross-codon forbidden IUPAC match under `strand_mode`. - Count changed nucleotides across the entire final CDS, including synonymous changes outside the codon for the requested substitution. ## Selection Minimize the total number of changed nucleotides permitted by this task; among equally minimal candidates, return the lexicographically smallest full DNA string (A<C<G<T). ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Expected answer `Final answer: ATGAAAGAGTTCTTCGTACTGTGA` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The A3E/I4F cloning substrate instead starts with Asp at position 3 (GAT). Its requested D3E edit must avoid introducing GAATTC. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Encode the requested amino-acid substitution and remove the forbidden DNA patterns using the fewest total nucleotide substitutions. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "residue_index": {{residue_index}}, "target_aa": "{{target_aa}}", "forbidden_patterns": {{forbidden_patterns}}, "strand_mode": "{{strand_mode}}" } ``` ## Rules - Change only the requested amino acid in the protein. Keep every other residue unchanged under NCBI genetic code 11. - Synonymous DNA substitutions at other codons are allowed when needed to remove forbidden patterns. Keep DNA length, the original first codon, and terminal stop unchanged. - Remove every overlapping or cross-codon forbidden IUPAC match under `strand_mode`. - Count changed nucleotides across the entire final CDS, including synonymous changes outside the codon for the requested substitution. ## Selection Minimize the total number of changed nucleotides permitted by this task; among equally minimal candidates, return the lexicographically smallest full DNA string (A<C<G<T). ## Answer - Return one continuous uppercase DNA sequence. If no valid result exists, return: ```text IMPOSSIBLE ``` After any explanation, write `Final answer:` on its own line. Put the required value or list on the next line to finish your response. Answer format and selection · scalar `output_format = scalar` Return one continuous uppercase DNA sequence. Selection Minimize the total number of changed nucleotides permitted by this task; among equally minimal candidates, return the lexicographically smallest full DNA string (A<C<G<T). If no valid result exists, return: IMPOSSIBLE Steps and final output Introduce the requested amino-acid substitution and use synonymous changes elsewhere to remove forbidden sites. Minimize the total number of nucleotide substitutions while meeting both requirements; break ties by DNA lexicographic order. The fewest edits needed for the amino-acid substitution alone provide a comparison, but those particular edits need not be retained in the final design. Output: canonical DNA string, or IMPOSSIBLE Literal task fields { "composition_type": "joint.protein_edit_and_cloning", "output_format": "scalar", "scorer": "exact_match", "mode": "joint", "selection_rule": "Minimize the total number of changed nucleotides permitted by this task; among equally minimal candidates, return the lexicographically smallest full DNA string (A<C<G<T)." } Reference material. Synonymous codons, NCBI genetic code 11: internal DNA codons, IUPAC DNA base sets, DNA reverse complement
Build a tagged construct, cleave it, measure products`pipeline.tag_cleave_mass`	Assemble a tagged CDS, translate it, simulate TEV cleavage, and calculate the mass of each resulting fragment.
Build a tagged construct, cleave it, measure products The input supplies tag and linker DNA plus its placement. Assemble the construct by the stated convention, translate the assembled DNA, and cut the product at the benchmark TEV sites. Then calculate each fragment’s mass separately. This tracks both module assembly and the residues left on each side of cleavage. Practical context This follows an affinity-tagged construct through tag removal to expected product identity. Keeping track of the exact cleavage scar matters because the purified target’s sequence and theoretical mass can differ from both the original untagged protein and the full fusion. Method references: Removal of affinity tags with TEV protease, ExPASy PeptideMass documentation. Component tasks Build a tagged protein CDS`construct.tag_fusion`Translate DNA to protein`cds.translate`Derive TEV-cleaved products`construct.tev_products`Calculate peptide mass`protein.monoisotopic_mass` Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `tag_cds`: `CAT CAT CAT CAT CAT CAT GAA AAT CTG TAT TTT CAG`; `linker_cds`: `GGT GGT TCT`; `placement`: `N`. Procedure. Build the same His6–TEV/GGS fusion as tag_fusion, translate it, and cut ENLYFQ\|G. Sum each fragment’s residue masses plus its own water molecule, rounding each fragment mass once. Final answer: { "cds": "ATGCATCATCATCATCATCATGAAAATCTGTATTTTCAGGGTGGTTCTAAAGCAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA", "fragments": [ { "peptide": "MHHHHHHENLYFQ", "mass_da": 1765.7644 }, { "peptide": "GGSKAIFVLKGWWRTS", "mass_da": 1791.9784 } ] } Checking the answer. Check the constructed DNA, all cleavage-product sequences, and each fragment’s calculated mass. Easy example · references supplied Prompt ## Task Build the tagged CDS, translate it, apply ENLYFQ\|G cleavage, and return the constructed DNA plus every resulting peptide and neutral monoisotopic mass. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "tag_cds": "CAT CAT CAT CAT CAT CAT GAA AAT CTG TAT TTT CAG", "linker_cds": "GGT GGT TCT", "placement": "N" } ``` ## Rules - For N placement, assemble: original first codon + tag + linker + remaining original sense codons + original stop. - For C placement, assemble: original sense codons + linker + tag + original stop. - Translate with NCBI genetic code 11 and first-stop termination. - Cut every exact `ENLYFQG` match as `ENLYFQ\|G`; G stays on the downstream fragment. - Calculate each unmodified fragment’s neutral monoisotopic mass independently, with free termini and one water per fragment. ## Mass constants Use these isotope masses in Da as exact decimal constants: ```text 12C = 12.00000000000 1H = 1.00782503223 14N = 14.00307400443 16O = 15.99491461957 32S = 31.9720711744 ``` For a neutral, unmodified linear peptide, sum the neutral free-amino-acid molecular formulas and subtract one H2O for each peptide bond. Equivalently, sum the incorporated residue masses and add one H2O for the complete peptide. Do not round intermediate values. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Monoisotopic residue masses Use these incorporated residue masses in Da as exact decimal constants. Each is the corresponding neutral free amino acid minus one H2O. ```text A=71.03711378515 C=103.00918495955 D=115.02694302429 E=129.04259308875 F=147.06841391407 G=57.02146372069 H=137.05891185847 I=113.08406397853 K=128.09496301519 L=113.08406397853 M=131.04048508847 N=114.04292744138 P=97.05276384961 Q=128.05857750584 R=156.10111102405 S=87.03202840472 T=101.04767846918 V=99.06841391407 W=186.07931295073 Y=163.06332853364 ``` Add one H2O = 18.01056468403 Da for the complete free linear peptide. ## Answer - `cds` is the exact constructed continuous uppercase DNA sequence. `fragments` lists every nonempty final TEV fragment in N-to-C order, retaining repeats. - `peptide` is its continuous uppercase amino-acid sequence. `mass_da` is its neutral monoisotopic mass in Da as a JSON number, rounded once to 4 decimal places with one water per fragment. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "cds": <DNA string>, "fragments": [ { "peptide": <amino-acid string>, "mass_da": <number> } ] } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "cds": "ATGCATCATCATCATCATCATGAAAATCTGTATTTTCAGGGTGGTTCTAAAGCAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA", "fragments": [ { "peptide": "MHHHHHHENLYFQ", "mass_da": 1765.7644 }, { "peptide": "GGSKAIFVLKGWWRTS", "mass_da": 1791.9784 } ] }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The supplied modules encode a His6 tag followed by a TEV site (HHHHHHENLYFQ) and a GGS linker. These are designed additions to native trpL. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Build the tagged CDS, translate it, apply ENLYFQ\|G cleavage, and return the constructed DNA plus every resulting peptide and neutral monoisotopic mass. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "tag_cds": "{{tag_cds}}", "linker_cds": "{{linker_cds}}", "placement": "{{placement}}" } ``` ## Rules - For N placement, assemble: original first codon + tag + linker + remaining original sense codons + original stop. - For C placement, assemble: original sense codons + linker + tag + original stop. - Translate with NCBI genetic code 11 and first-stop termination. - Cut every exact `ENLYFQG` match as `ENLYFQ\|G`; G stays on the downstream fragment. - Calculate each unmodified fragment’s neutral monoisotopic mass independently, with free termini and one water per fragment. ## Mass constants Use these isotope masses in Da as exact decimal constants: ```text 12C = 12.00000000000 1H = 1.00782503223 14N = 14.00307400443 16O = 15.99491461957 32S = 31.9720711744 ``` For a neutral, unmodified linear peptide, sum the neutral free-amino-acid molecular formulas and subtract one H2O for each peptide bond. Equivalently, sum the incorporated residue masses and add one H2O for the complete peptide. Do not round intermediate values. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Monoisotopic residue masses Use these incorporated residue masses in Da as exact decimal constants. Each is the corresponding neutral free amino acid minus one H2O. ```text A=71.03711378515 C=103.00918495955 D=115.02694302429 E=129.04259308875 F=147.06841391407 G=57.02146372069 H=137.05891185847 I=113.08406397853 K=128.09496301519 L=113.08406397853 M=131.04048508847 N=114.04292744138 P=97.05276384961 Q=128.05857750584 R=156.10111102405 S=87.03202840472 T=101.04767846918 V=99.06841391407 W=186.07931295073 Y=163.06332853364 ``` Add one H2O = 18.01056468403 Da for the complete free linear peptide. ## Answer - `cds` is the exact constructed continuous uppercase DNA sequence. `fragments` lists every nonempty final TEV fragment in N-to-C order, retaining repeats. - `peptide` is its continuous uppercase amino-acid sequence. `mass_da` is its neutral monoisotopic mass in Da as a JSON number, rounded once to 4 decimal places with one water per fragment. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "cds": <DNA string>, "fragments": [ { "peptide": <amino-acid string>, "mass_da": <number> } ] } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Build the tagged CDS, translate it, apply ENLYFQ\|G cleavage, and return the constructed DNA plus every resulting peptide and neutral monoisotopic mass. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "tag_cds": "CAT CAT CAT CAT CAT CAT GAA AAT CTG TAT TTT CAG", "linker_cds": "GGT GGT TCT", "placement": "N" } ``` ## Rules - For N placement, assemble: original first codon + tag + linker + remaining original sense codons + original stop. - For C placement, assemble: original sense codons + linker + tag + original stop. - Translate with NCBI genetic code 11 and first-stop termination. - Cut every exact `ENLYFQG` match as `ENLYFQ\|G`; G stays on the downstream fragment. - Calculate each unmodified fragment’s neutral monoisotopic mass independently, with free termini and one water per fragment. ## Mass constants Use these isotope masses in Da as exact decimal constants: ```text 12C = 12.00000000000 1H = 1.00782503223 14N = 14.00307400443 16O = 15.99491461957 32S = 31.9720711744 ``` For a neutral, unmodified linear peptide, sum the neutral free-amino-acid molecular formulas and subtract one H2O for each peptide bond. Equivalently, sum the incorporated residue masses and add one H2O for the complete peptide. Do not round intermediate values. ## Answer - `cds` is the exact constructed continuous uppercase DNA sequence. `fragments` lists every nonempty final TEV fragment in N-to-C order, retaining repeats. - `peptide` is its continuous uppercase amino-acid sequence. `mass_da` is its neutral monoisotopic mass in Da as a JSON number, rounded once to 4 decimal places with one water per fragment. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "cds": <DNA string>, "fragments": [ { "peptide": <amino-acid string>, "mass_da": <number> } ] } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "cds": "ATGCATCATCATCATCATCATGAAAATCTGTATTTTCAGGGTGGTTCTAAAGCAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA", "fragments": [ { "peptide": "MHHHHHHENLYFQ", "mass_da": 1765.7644 }, { "peptide": "GGSKAIFVLKGWWRTS", "mass_da": 1791.9784 } ] }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The supplied modules encode a His6 tag followed by a TEV site (HHHHHHENLYFQ) and a GGS linker. These are designed additions to native trpL. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Build the tagged CDS, translate it, apply ENLYFQ\|G cleavage, and return the constructed DNA plus every resulting peptide and neutral monoisotopic mass. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "tag_cds": "{{tag_cds}}", "linker_cds": "{{linker_cds}}", "placement": "{{placement}}" } ``` ## Rules - For N placement, assemble: original first codon + tag + linker + remaining original sense codons + original stop. - For C placement, assemble: original sense codons + linker + tag + original stop. - Translate with NCBI genetic code 11 and first-stop termination. - Cut every exact `ENLYFQG` match as `ENLYFQ\|G`; G stays on the downstream fragment. - Calculate each unmodified fragment’s neutral monoisotopic mass independently, with free termini and one water per fragment. ## Mass constants Use these isotope masses in Da as exact decimal constants: ```text 12C = 12.00000000000 1H = 1.00782503223 14N = 14.00307400443 16O = 15.99491461957 32S = 31.9720711744 ``` For a neutral, unmodified linear peptide, sum the neutral free-amino-acid molecular formulas and subtract one H2O for each peptide bond. Equivalently, sum the incorporated residue masses and add one H2O for the complete peptide. Do not round intermediate values. ## Answer - `cds` is the exact constructed continuous uppercase DNA sequence. `fragments` lists every nonempty final TEV fragment in N-to-C order, retaining repeats. - `peptide` is its continuous uppercase amino-acid sequence. `mass_da` is its neutral monoisotopic mass in Da as a JSON number, rounded once to 4 decimal places with one water per fragment. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "cds": <DNA string>, "fragments": [ { "peptide": <amino-acid string>, "mass_da": <number> } ] } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `cds` is the exact constructed continuous uppercase DNA sequence. `fragments` lists every nonempty final TEV fragment in N-to-C order, retaining repeats. `peptide` is its continuous uppercase amino-acid sequence. `mass_da` is its neutral monoisotopic mass in Da as a JSON number, rounded once to 4 decimal places with one water per fragment. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "cds": <DNA string>, "fragments": [ { "peptide": <amino-acid string>, "mass_da": <number> } ] } Steps and final output Construct the tagged CDS, translate it, and apply the TEV cleavage rule. Calculate the monoisotopic mass of each final fragment, adding one H2O per fragment. The supplied tag and linker must encode the intended TEV site; residues remaining on each side follow the ENLYFQ\|G cut. Output: JSON object with exactly cds, fragments Literal task fields { "composition_type": "pipeline.tag_cleave_mass", "output_format": "json", "scorer": "exact_match", "mode": "pipeline", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Reference material. Annotated CDS initiation, NCBI genetic code 11: internal DNA codons, Monoisotopic isotope masses, Monoisotopic residue masses
Design a codon library and verify its coverage`pipeline.library_and_verify`	Design a degenerate codon for requested amino acids, then expand it to verify coverage and absence of stops.
Design a codon library and verify its coverage Choose the IUPAC codon under the focused-library task’s optimization rules. Insert that notation at the selected internal CDS position, enumerate every concrete codon, and count its protein products. The final enumeration checks what the compact notation actually permits rather than relying on the intended target list. Practical context This connects the intent of a focused mutagenesis library to what will actually be synthesized. Enumerating the proposed degenerate codon checks target amino-acid coverage, unintended products, and stops before ordering the oligonucleotide mixture. Method reference: SwiftLib. Component tasks Design a focused degenerate codon`library.design_codon`Enumerate a degenerate codon library`library.codon_products`Assess an ambiguous stop`ambiguity.stop_certainty` Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `codon_index`: `3`; `target_residues`: `["A", "G", "V"]`. Procedure. Choose the optimal stop-free IUPAC triplet for A/G/V coverage, then expand that selected triplet again and count each concrete codon’s internal amino-acid product. Report the design and its complete expansion counts. Final answer: { "status": "ok", "codon": "GBA", "residue_counts": { "A": 1, "G": 1, "V": 1 } } Checking the answer. The chosen codon must be optimal under the declared objective, and its reported residue multiplicities must match its full expansion. Easy example · references supplied Prompt ## Task Design a stop-free IUPAC codon covering every target amino acid, then report how many concrete codons encode each amino acid in its expansion. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "codon_index": 3, "target_residues": ["A", "G", "V"] } ``` ## Rules - Choose a single IUPAC triplet for the selected internal codon. Its expansions must cover all `target_residues` and contain no NCBI genetic code 11 stop. - Expand that same selected triplet into distinct concrete DNA codons, translate each with its ordinary internal meaning, and count how many encode each amino acid. ## Selection Require target-residue coverage and no stops. Minimize extra amino-acid types, then the number of concrete codons, then the IUPAC triplet in ASCII order A<B<C<D<G<H<K<M<N<R<S<T<V<W<Y. Compare the symbols themselves, not their expanded nucleotide sets. ## Reference material ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - `codon` is the selected uppercase IUPAC DNA triplet. `residue_counts` maps every produced one-letter amino-acid symbol to the number of distinct concrete codons encoding it in that triplet’s expansion. - Include all positive counts, including amino acids outside target_residues. Their sum is the expansion size. These are codon multiplicities rather than probabilities or full-protein residue counts. A valid design is stop-free. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "codon": <IUPAC DNA triplet>, "residue_counts": {"<one-letter amino-acid symbol>": <positive integer count>} } ``` If no valid result exists, return: ```json { "status": "impossible", "codon": null, "residue_counts": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "status": "ok", "codon": "GBA", "residue_counts": { "A": 1, "G": 1, "V": 1 } }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Design a stop-free IUPAC codon covering every target amino acid, then report how many concrete codons encode each amino acid in its expansion. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "codon_index": {{codon_index}}, "target_residues": {{target_residues}} } ``` ## Rules - Choose a single IUPAC triplet for the selected internal codon. Its expansions must cover all `target_residues` and contain no NCBI genetic code 11 stop. - Expand that same selected triplet into distinct concrete DNA codons, translate each with its ordinary internal meaning, and count how many encode each amino acid. ## Selection Require target-residue coverage and no stops. Minimize extra amino-acid types, then the number of concrete codons, then the IUPAC triplet in ASCII order A<B<C<D<G<H<K<M<N<R<S<T<V<W<Y. Compare the symbols themselves, not their expanded nucleotide sets. ## Reference material ### IUPAC DNA base sets An IUPAC symbol permits any base in its listed set. Choices at different positions are independent. ```text A=A C=C G=G T=T R=AG Y=CT S=CG W=AT K=GT M=AC B=CGT D=AGT H=ACT V=ACG N=ACGT ``` ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - `codon` is the selected uppercase IUPAC DNA triplet. `residue_counts` maps every produced one-letter amino-acid symbol to the number of distinct concrete codons encoding it in that triplet’s expansion. - Include all positive counts, including amino acids outside target_residues. Their sum is the expansion size. These are codon multiplicities rather than probabilities or full-protein residue counts. A valid design is stop-free. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "codon": <IUPAC DNA triplet>, "residue_counts": {"<one-letter amino-acid symbol>": <positive integer count>} } ``` If no valid result exists, return: ```json { "status": "impossible", "codon": null, "residue_counts": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Design a stop-free IUPAC codon covering every target amino acid, then report how many concrete codons encode each amino acid in its expansion. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "codon_index": 3, "target_residues": ["A", "G", "V"] } ``` ## Rules - Choose a single IUPAC triplet for the selected internal codon. Its expansions must cover all `target_residues` and contain no NCBI genetic code 11 stop. - Expand that same selected triplet into distinct concrete DNA codons, translate each with its ordinary internal meaning, and count how many encode each amino acid. ## Selection Require target-residue coverage and no stops. Minimize extra amino-acid types, then the number of concrete codons, then the IUPAC triplet in ASCII order A<B<C<D<G<H<K<M<N<R<S<T<V<W<Y. Compare the symbols themselves, not their expanded nucleotide sets. ## Answer - `codon` is the selected uppercase IUPAC DNA triplet. `residue_counts` maps every produced one-letter amino-acid symbol to the number of distinct concrete codons encoding it in that triplet’s expansion. - Include all positive counts, including amino acids outside target_residues. Their sum is the expansion size. These are codon multiplicities rather than probabilities or full-protein residue counts. A valid design is stop-free. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "codon": <IUPAC DNA triplet>, "residue_counts": {"<one-letter amino-acid symbol>": <positive integer count>} } ``` If no valid result exists, return: ```json { "status": "impossible", "codon": null, "residue_counts": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "status": "ok", "codon": "GBA", "residue_counts": { "A": 1, "G": 1, "V": 1 } }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Design a stop-free IUPAC codon covering every target amino acid, then report how many concrete codons encode each amino acid in its expansion. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds": "{{cds}}", "codon_index": {{codon_index}}, "target_residues": {{target_residues}} } ``` ## Rules - Choose a single IUPAC triplet for the selected internal codon. Its expansions must cover all `target_residues` and contain no NCBI genetic code 11 stop. - Expand that same selected triplet into distinct concrete DNA codons, translate each with its ordinary internal meaning, and count how many encode each amino acid. ## Selection Require target-residue coverage and no stops. Minimize extra amino-acid types, then the number of concrete codons, then the IUPAC triplet in ASCII order A<B<C<D<G<H<K<M<N<R<S<T<V<W<Y. Compare the symbols themselves, not their expanded nucleotide sets. ## Answer - `codon` is the selected uppercase IUPAC DNA triplet. `residue_counts` maps every produced one-letter amino-acid symbol to the number of distinct concrete codons encoding it in that triplet’s expansion. - Include all positive counts, including amino acids outside target_residues. Their sum is the expansion size. These are codon multiplicities rather than probabilities or full-protein residue counts. A valid design is stop-free. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "codon": <IUPAC DNA triplet>, "residue_counts": {"<one-letter amino-acid symbol>": <positive integer count>} } ``` If no valid result exists, return: ```json { "status": "impossible", "codon": null, "residue_counts": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `codon` is the selected uppercase IUPAC DNA triplet. `residue_counts` maps every produced one-letter amino-acid symbol to the number of distinct concrete codons encoding it in that triplet’s expansion. Include all positive counts, including amino acids outside target_residues. Their sum is the expansion size. These are codon multiplicities rather than probabilities or full-protein residue counts. A valid design is stop-free. Selection Require target-residue coverage and no stops. Minimize extra amino-acid types, then the number of concrete codons, then the IUPAC triplet in ASCII order A<B<C<D<G<H<K<M<N<R<S<T<V<W<Y. Compare the symbols themselves, not their expanded nucleotide sets. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "status": "ok", "codon": <IUPAC DNA triplet>, "residue_counts": {"<one-letter amino-acid symbol>": <positive integer count>} } If no valid result exists, return: { "status": "impossible", "codon": null, "residue_counts": null } Steps and final output Replace the selected codon with the chosen IUPAC triplet. Expand that same codon to enumerate its amino-acid products and check whether it can encode a stop. The products must cover target_residues, and none of the codon expansions may be a stop (stop status=never). Output: JSON object with exactly status, codon, residue_counts Literal task fields { "composition_type": "pipeline.library_and_verify", "output_format": "json", "scorer": "exact_match", "mode": "pipeline", "selection_rule": "Require target-residue coverage and no stops. Minimize extra amino-acid types, then the number of concrete codons, then the IUPAC triplet in ASCII order A<B<C<D<G<H<K<M<N<R<S<T<V<W<Y. Compare the symbols themselves, not their expanded nucleotide sets." } Reference material. IUPAC DNA base sets, NCBI genetic code 11: internal DNA codons
Assemble coding modules and locate junction motifs`pipeline.fusion_junction`	Join two CDS modules, find peptide-pattern matches in the fusion product, and map those matches to the fused DNA.
Assemble coding modules and locate junction motifs Assemble the modules using the supplied linker and start/stop convention, then translate the fusion. Search the whole protein, including patterns formed across a module junction. Map each result to coordinates in the new fused CDS, not to the coordinate systems of the two original inputs. Practical context A fusion can introduce a peptide feature that neither module contains on its own. Checking across each junction and mapping a hit back to DNA supplies the information needed for a subsequent allowed repair, such as removing an unwanted protease site. Component tasks Join coding modules`construct.join_modules`Translate DNA to protein`cds.translate`Find a supplied peptide pattern`protein.motif_positions`Map residues back to DNA`coords.aa_to_nt` Task Outline Inputs. `cds_a`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `cds_b`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `linker_cds`: `GGT GGT TCT`; `pattern`: `SGGSK`. Procedure. Build the same full-length/GGS/minigene fusion as join_modules. SGGSK crosses from the first module’s final Ser through the linker to the second module’s Lys; locate it in the fused protein and map to fused-DNA coordinates. Final answer: { "fused_cds": "ATGAAAGCAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCGGTGGTTCTAAAGCAATTTTCGTACTGTGA", "matches": [ { "aa_interval": [14, 18], "nt_interval": [40, 54] } ] } Checking the answer. The fused DNA, all protein matches, and their fused-DNA intervals must be correct. Easy example · references supplied Prompt ## Task Fuse A–linker–B under the module-join rule, find every peptide-pattern match, and report its residue and fused-DNA intervals. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds_a": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "cds_b": "ATG AAA GCA ATT TTC GTA CTG TGA", "linker_cds": "GGT GGT TCT", "pattern": "SGGSK" } ``` ## Rules - Remove A’s terminal stop and B’s first codon, then concatenate the remaining A DNA, linker, and remaining B DNA. Keep B’s terminal stop. - Translate the entire fused CDS with NCBI genetic code 11, ending at its first in-frame stop. - Find every overlapping peptide-pattern match, including matches spanning the assembly junction. - Pattern syntax: a letter matches itself; `X` matches any residue; `[set]` matches a listed residue; `{set}` excludes listed residues. - Report DNA intervals in the assembled CDS’s coordinates, not in either source CDS’s coordinates. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - `fused_cds` is the exact assembled continuous uppercase DNA sequence. - Each match gives aa_interval=[first_residue, last_residue] in the fused protein and nt_interval=[first_base, last_base] in fused_cds, both with 1-based inclusive integer endpoints. - Include all overlapping matches, sorted by protein start and then end. An empty match list is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "fused_cds": <DNA string>, "matches": [ { "aa_interval": [<first residue>, <last residue>], "nt_interval": [<first base>, <last base>] } ] } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "fused_cds": "ATGAAAGCAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCGGTGGTTCTAAAGCAATTTTCGTACTGTGA", "matches": [ { "aa_interval": [14, 18], "nt_interval": [40, 54] } ] }` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Join native full-length trpL to the explicitly truncated trpL minigene using a designed GGS linker. This is a constructed tandem fusion, not a natural gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Fuse A–linker–B under the module-join rule, find every peptide-pattern match, and report its residue and fused-DNA intervals. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds_a": "{{cds_a}}", "cds_b": "{{cds_b}}", "linker_cds": "{{linker_cds}}", "pattern": "{{pattern}}" } ``` ## Rules - Remove A’s terminal stop and B’s first codon, then concatenate the remaining A DNA, linker, and remaining B DNA. Keep B’s terminal stop. - Translate the entire fused CDS with NCBI genetic code 11, ending at its first in-frame stop. - Find every overlapping peptide-pattern match, including matches spanning the assembly junction. - Pattern syntax: a letter matches itself; `X` matches any residue; `[set]` matches a listed residue; `{set}` excludes listed residues. - Report DNA intervals in the assembled CDS’s coordinates, not in either source CDS’s coordinates. ## Reference material ### Annotated CDS initiation The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - `fused_cds` is the exact assembled continuous uppercase DNA sequence. - Each match gives aa_interval=[first_residue, last_residue] in the fused protein and nt_interval=[first_base, last_base] in fused_cds, both with 1-based inclusive integer endpoints. - Include all overlapping matches, sorted by protein start and then end. An empty match list is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "fused_cds": <DNA string>, "matches": [ { "aa_interval": [<first residue>, <last residue>], "nt_interval": [<first base>, <last base>] } ] } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Fuse A–linker–B under the module-join rule, find every peptide-pattern match, and report its residue and fused-DNA intervals. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds_a": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "cds_b": "ATG AAA GCA ATT TTC GTA CTG TGA", "linker_cds": "GGT GGT TCT", "pattern": "SGGSK" } ``` ## Rules - Remove A’s terminal stop and B’s first codon, then concatenate the remaining A DNA, linker, and remaining B DNA. Keep B’s terminal stop. - Translate the entire fused CDS with NCBI genetic code 11, ending at its first in-frame stop. - Find every overlapping peptide-pattern match, including matches spanning the assembly junction. - Pattern syntax: a letter matches itself; `X` matches any residue; `[set]` matches a listed residue; `{set}` excludes listed residues. - Report DNA intervals in the assembled CDS’s coordinates, not in either source CDS’s coordinates. ## Answer - `fused_cds` is the exact assembled continuous uppercase DNA sequence. - Each match gives aa_interval=[first_residue, last_residue] in the fused protein and nt_interval=[first_base, last_base] in fused_cds, both with 1-based inclusive integer endpoints. - Include all overlapping matches, sorted by protein start and then end. An empty match list is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "fused_cds": <DNA string>, "matches": [ { "aa_interval": [<first residue>, <last residue>], "nt_interval": [<first base>, <last base>] } ] } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "fused_cds": "ATGAAAGCAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCGGTGGTTCTAAAGCAATTTTCGTACTGTGA", "matches": [ { "aa_interval": [14, 18], "nt_interval": [40, 54] } ] }` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Join native full-length trpL to the explicitly truncated trpL minigene using a designed GGS linker. This is a constructed tandem fusion, not a natural gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Fuse A–linker–B under the module-join rule, find every peptide-pattern match, and report its residue and fused-DNA intervals. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. ```json { "cds_a": "{{cds_a}}", "cds_b": "{{cds_b}}", "linker_cds": "{{linker_cds}}", "pattern": "{{pattern}}" } ``` ## Rules - Remove A’s terminal stop and B’s first codon, then concatenate the remaining A DNA, linker, and remaining B DNA. Keep B’s terminal stop. - Translate the entire fused CDS with NCBI genetic code 11, ending at its first in-frame stop. - Find every overlapping peptide-pattern match, including matches spanning the assembly junction. - Pattern syntax: a letter matches itself; `X` matches any residue; `[set]` matches a listed residue; `{set}` excludes listed residues. - Report DNA intervals in the assembled CDS’s coordinates, not in either source CDS’s coordinates. ## Answer - `fused_cds` is the exact assembled continuous uppercase DNA sequence. - Each match gives aa_interval=[first_residue, last_residue] in the fused protein and nt_interval=[first_base, last_base] in fused_cds, both with 1-based inclusive integer endpoints. - Include all overlapping matches, sorted by protein start and then end. An empty match list is []. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "fused_cds": <DNA string>, "matches": [ { "aa_interval": [<first residue>, <last residue>], "nt_interval": [<first base>, <last base>] } ] } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `fused_cds` is the exact assembled continuous uppercase DNA sequence. Each match gives aa_interval=[first_residue, last_residue] in the fused protein and nt_interval=[first_base, last_base] in fused_cds, both with 1-based inclusive integer endpoints. Include all overlapping matches, sorted by protein start and then end. An empty match list is []. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "fused_cds": <DNA string>, "matches": [ { "aa_interval": [<first residue>, <last residue>], "nt_interval": [<first base>, <last base>] } ] } Steps and final output Translate the fused CDS and search the entire resulting protein, including motifs spanning the junction. Convert each matching peptide interval into coordinates in the fused DNA sequence. Output: JSON object with exactly fused_cds, matches Literal task fields { "composition_type": "pipeline.fusion_junction", "output_format": "json", "scorer": "exact_match", "mode": "pipeline", "selection_rule": "Apply the stated procedure; the result is uniquely determined by the inputs." } Reference material. Annotated CDS initiation, NCBI genetic code 11: internal DNA codons
Design and verify an RNA-editing-dependent product`pipeline.edit_dependent_product`	Design a CDS that preserves the original protein before RNA editing and gives a specified protein after the supplied edits.
Design and verify an RNA-editing-dependent product First find a permitted synonymous encoding satisfying the editing-dependent design task. Then verify two branches from that same candidate: ordinary translation must produce the original protein, and transcription followed by the supplied RNA edits and translation must produce the target protein. Returning both proteins makes the two requirements inspectable. Exercise scope ADAR-mediated RNA recoding is the biological inspiration for the revised A→G translation example. The two branches check the unedited protein and the product implied by the supplied edit. This audits encoding logic; it does not design an ADAR recognition site, guide, or editing reaction. Method reference: Programmable editing by endogenous ADAR. Component tasks Design an RNA-editing-dependent product`construct.edit_aware_encoding`Translate DNA to protein`cds.translate`Make the coding RNA`dna.transcribe`Apply specified RNA edits`rna.apply_edits`Translate coding RNA`rna.translate` Task Outline Inputs. `cds`: `ATG AAA TCT ATT TTC GTA CTG TGA`; `rna_edits`: `[{"replace": [7, 7, "G"]}]`; `target_protein`: `MKGIFVL`; `editable_positions`: `[7, 8]`. Procedure. Choose Ser codon AGT at position 3 while preserving MKSIFVL before editing. The supplied G substitution at RNA base 7 yields Gly codon GGU; report the final DNA and both complete protein sequences. Final answer: { "status": "ok", "cds": "ATGAAAAGTATTTTCGTACTGTGA", "unedited_protein": "MKSIFVL", "edited_protein": "MKGIFVL" } Checking the answer. Verify the DNA permissions and both exact protein products on the same designed sequence. Easy example · references supplied Prompt ## Task Design a synonymous CDS that produces the original protein before RNA editing and the requested protein afterward. Report the DNA and both protein sequences. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA TCT ATT TTC GTA CTG TGA", "rna_edits": [ { "replace": [7, 7, "G"] } ], "target_protein": "MKGIFVL", "editable_positions": [7, 8] } ``` ## Rules - Only DNA positions in `editable_positions` may change. Keep length, the original first codon, and terminal stop unchanged. - Before RNA editing, the candidate must encode the complete original protein under NCBI genetic code 11. - Transcribe that same candidate and apply exactly the supplied edits using original RNA coordinates and RNA replacement text. - Translate the edited RNA from its original reading origin to the first in-frame stop. Omit the stop symbol. The resulting protein must equal `target_protein`. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Initiation of coding RNA At the first RNA codon, UUG, CUG, AUU, AUC, AUA, AUG, and GUG encode initiating M under NCBI genetic code 11. If that first codon is not an allowed initiator, use its ordinary internal assignment. ### NCBI genetic code 11: internal RNA codons Each internal RNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text UUU=F UUC=F UUA=L UUG=L UCU=S UCC=S UCA=S UCG=S UAU=Y UAC=Y UAA=STOP UAG=STOP UGU=C UGC=C UGA=STOP UGG=W CUU=L CUC=L CUA=L CUG=L CCU=P CCC=P CCA=P CCG=P CAU=H CAC=H CAA=Q CAG=Q CGU=R CGC=R CGA=R CGG=R AUU=I AUC=I AUA=I AUG=M ACU=T ACC=T ACA=T ACG=T AAU=N AAC=N AAA=K AAG=K AGU=S AGC=S AGA=R AGG=R GUU=V GUC=V GUA=V GUG=V GCU=A GCC=A GCA=A GCG=A GAU=D GAC=D GAA=E GAG=E GGU=G GGC=G GGA=G GGG=G ``` ## Answer - `cds` is the selected continuous uppercase DNA sequence. `unedited_protein` is its translation before RNA editing; `edited_protein` is its translation after transcription and the supplied RNA edits. - Protein strings are continuous uppercase one-letter amino-acid sequences, ending before the first in-frame stop. A zero-residue protein is the empty JSON string "". Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "cds": <DNA string>, "unedited_protein": <amino-acid string>, "edited_protein": <amino-acid string> } ``` If no valid result exists, return: ```json { "status": "impossible", "cds": null, "unedited_protein": null, "edited_protein": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "status": "ok", "cds": "ATGAAAAGTATTTTCGTACTGTGA", "unedited_protein": "MKSIFVL", "edited_protein": "MKGIFVL" }` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The minigene has A3S (GCA→TCT). The supplied RNA edit sets base 7 to G; choose a synonymous Ser encoding so that this edit yields Gly. No naturally observed editing is claimed. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Design a synonymous CDS that produces the original protein before RNA editing and the requested protein afterward. Report the DNA and both protein sequences. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "rna_edits": {{rna_edits}}, "target_protein": "{{target_protein}}", "editable_positions": {{editable_positions}} } ``` ## Rules - Only DNA positions in `editable_positions` may change. Keep length, the original first codon, and terminal stop unchanged. - Before RNA editing, the candidate must encode the complete original protein under NCBI genetic code 11. - Transcribe that same candidate and apply exactly the supplied edits using original RNA coordinates and RNA replacement text. - Translate the edited RNA from its original reading origin to the first in-frame stop. Omit the stop symbol. The resulting protein must equal `target_protein`. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### Initiation of coding RNA At the first RNA codon, UUG, CUG, AUU, AUC, AUA, AUG, and GUG encode initiating M under NCBI genetic code 11. If that first codon is not an allowed initiator, use its ordinary internal assignment. ### NCBI genetic code 11: internal RNA codons Each internal RNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text UUU=F UUC=F UUA=L UUG=L UCU=S UCC=S UCA=S UCG=S UAU=Y UAC=Y UAA=STOP UAG=STOP UGU=C UGC=C UGA=STOP UGG=W CUU=L CUC=L CUA=L CUG=L CCU=P CCC=P CCA=P CCG=P CAU=H CAC=H CAA=Q CAG=Q CGU=R CGC=R CGA=R CGG=R AUU=I AUC=I AUA=I AUG=M ACU=T ACC=T ACA=T ACG=T AAU=N AAC=N AAA=K AAG=K AGU=S AGC=S AGA=R AGG=R GUU=V GUC=V GUA=V GUG=V GCU=A GCC=A GCA=A GCG=A GAU=D GAC=D GAA=E GAG=E GGU=G GGC=G GGA=G GGG=G ``` ## Answer - `cds` is the selected continuous uppercase DNA sequence. `unedited_protein` is its translation before RNA editing; `edited_protein` is its translation after transcription and the supplied RNA edits. - Protein strings are continuous uppercase one-letter amino-acid sequences, ending before the first in-frame stop. A zero-residue protein is the empty JSON string "". Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "cds": <DNA string>, "unedited_protein": <amino-acid string>, "edited_protein": <amino-acid string> } ``` If no valid result exists, return: ```json { "status": "impossible", "cds": null, "unedited_protein": null, "edited_protein": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Design a synonymous CDS that produces the original protein before RNA editing and the requested protein afterward. Report the DNA and both protein sequences. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "ATG AAA TCT ATT TTC GTA CTG TGA", "rna_edits": [ { "replace": [7, 7, "G"] } ], "target_protein": "MKGIFVL", "editable_positions": [7, 8] } ``` ## Rules - Only DNA positions in `editable_positions` may change. Keep length, the original first codon, and terminal stop unchanged. - Before RNA editing, the candidate must encode the complete original protein under NCBI genetic code 11. - Transcribe that same candidate and apply exactly the supplied edits using original RNA coordinates and RNA replacement text. - Translate the edited RNA from its original reading origin to the first in-frame stop. Omit the stop symbol. The resulting protein must equal `target_protein`. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - `cds` is the selected continuous uppercase DNA sequence. `unedited_protein` is its translation before RNA editing; `edited_protein` is its translation after transcription and the supplied RNA edits. - Protein strings are continuous uppercase one-letter amino-acid sequences, ending before the first in-frame stop. A zero-residue protein is the empty JSON string "". Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "cds": <DNA string>, "unedited_protein": <amino-acid string>, "edited_protein": <amino-acid string> } ``` If no valid result exists, return: ```json { "status": "impossible", "cds": null, "unedited_protein": null, "edited_protein": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "status": "ok", "cds": "ATGAAAAGTATTTTCGTACTGTGA", "unedited_protein": "MKSIFVL", "edited_protein": "MKGIFVL" }` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. The minigene has A3S (GCA→TCT). The supplied RNA edit sets base 7 to G; choose a synonymous Ser encoding so that this edit yields Gly. No naturally observed editing is claimed. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Design a synonymous CDS that produces the original protein before RNA editing and the requested protein afterward. Report the DNA and both protein sequences. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. In an edit object, `replace` is `[first_base, last_base, new_text]`; an empty `new_text` deletes that interval. `insert_after` gives an original base boundary and `sequence` gives the inserted text. Apply all edits to the original coordinates, simultaneously. ```json { "cds": "{{cds}}", "rna_edits": {{rna_edits}}, "target_protein": "{{target_protein}}", "editable_positions": {{editable_positions}} } ``` ## Rules - Only DNA positions in `editable_positions` may change. Keep length, the original first codon, and terminal stop unchanged. - Before RNA editing, the candidate must encode the complete original protein under NCBI genetic code 11. - Transcribe that same candidate and apply exactly the supplied edits using original RNA coordinates and RNA replacement text. - Translate the edited RNA from its original reading origin to the first in-frame stop. Omit the stop symbol. The resulting protein must equal `target_protein`. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - `cds` is the selected continuous uppercase DNA sequence. `unedited_protein` is its translation before RNA editing; `edited_protein` is its translation after transcription and the supplied RNA edits. - Protein strings are continuous uppercase one-letter amino-acid sequences, ending before the first in-frame stop. A zero-residue protein is the empty JSON string "". Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "cds": <DNA string>, "unedited_protein": <amino-acid string>, "edited_protein": <amino-acid string> } ``` If no valid result exists, return: ```json { "status": "impossible", "cds": null, "unedited_protein": null, "edited_protein": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `cds` is the selected continuous uppercase DNA sequence. `unedited_protein` is its translation before RNA editing; `edited_protein` is its translation after transcription and the supplied RNA edits. Protein strings are continuous uppercase one-letter amino-acid sequences, ending before the first in-frame stop. A zero-residue protein is the empty JSON string "". Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "status": "ok", "cds": <DNA string>, "unedited_protein": <amino-acid string>, "edited_protein": <amino-acid string> } If no valid result exists, return: { "status": "impossible", "cds": null, "unedited_protein": null, "edited_protein": null } Steps and final output Translate the candidate CDS unchanged and verify that it encodes the original protein. Separately, transcribe the same candidate, apply the exact supplied RNA edits, and translate the edited RNA to verify target_protein. Apply the RNA edits to RNA, using its U alphabet. Output: JSON object with exactly status, cds, unedited_protein, edited_protein Literal task fields { "composition_type": "pipeline.edit_dependent_product", "output_format": "json", "scorer": "exact_match", "mode": "branch", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." } Reference material. Synonymous codons, Coding DNA to RNA, Initiation of coding RNA, NCBI genetic code 11: internal RNA codons
Edit one frame and audit both products`pipeline.overlap_edit_verify`	Design a change for one annotated coding window while preserving the overlapping product, then report both before-and-after peptides.
Edit one frame and audit both products Use the supplied windows, strands, target peptide, and edit permissions to design DNA. Compare original and designed DNA by translating each window under the overlap task’s internal-codon convention. The explicit before/after output checks both the intended change and the product that must stay fixed. Practical context This is a construct audit for a known coding overlap. Reporting both original and edited products verifies that the desired peptide change was achieved without silently altering the protected product. Its scope is sequence preservation under supplied annotations, rather than functional design of new overlapping genes. Component tasks Change one frame, preserve the other`overlap.constrained_edit`Evaluate a change in two frames`overlap.compare_products` Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA`; `frames`: `[{"interval": [7, 12], "strand": "+"}, {"interval": [6, 11], "strand": "+"}]`; `target_peptide_a`: `VI`; `editable_positions`: `[8]`. Procedure. Apply the C8T solution, then independently translate both fixed overlapping windows before and after. Frame A changes AI→VI; frame B stays SN→SN. Final answer: { "status": "ok", "cds": "ATGAAAGTAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA", "frames": [ { "frame": "A", "original_peptide": "AI", "edited_peptide": "VI", "changed": true }, { "frame": "B", "original_peptide": "SN", "edited_peptide": "SN", "changed": false } ] } Checking the answer. Check the designed DNA, both pairs of peptide strings, and every edit restriction. Easy example · references supplied Prompt ## Task Use permitted DNA substitutions to change frame A’s peptide while preserving frame B’s peptide. Report the final DNA and both frames’ original and edited peptides. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `frames` lists frame A first and frame B second. Each record supplies its inclusive DNA interval and strand. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "frames": [ { "interval": [7, 12], "strand": "+" }, { "interval": [6, 11], "strand": "+" } ], "target_peptide_a": "VI", "editable_positions": [8] } ``` ## Rules - Only DNA positions in `editable_positions` may change, through equal-length substitutions. - Extract the same two annotated windows before and after the edit. Reverse-complement any minus-strand window. - Use ordinary internal NCBI genetic code 11 meanings for every triplet, with no initiation override. Keep stops as `` and continue translating after them. - The final A peptide must equal `target_peptide_a`. The complete B peptide must remain exactly unchanged. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - `cds` is the selected continuous uppercase DNA sequence. `frames` contains exactly two records, "A" then "B", in supplied window order. - `original_peptide` and `edited_peptide` are the full internal translations before and after the DNA change, retaining every stop as . `changed` is true exactly when those strings differ. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "cds": <DNA string>, "frames": [ { "frame": <"A" or "B">, "original_peptide": <amino-acid string, retaining stops as >, "edited_peptide": <amino-acid string, retaining stops as >, "changed": <boolean> } ] } ``` If no valid result exists, return: ```json { "status": "impossible", "cds": null, "frames": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "status": "ok", "cds": "ATGAAAGTAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA", "frames": [ { "frame": "A", "original_peptide": "AI", "edited_peptide": "VI", "changed": true }, { "frame": "B", "original_peptide": "SN", "edited_peptide": "SN", "changed": false } ] }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The two overlapping test windows are supplied annotations, not two natural genes. C8T changes frame A Ala→Val while preserving the overlapping frame B Ser codon AGC→AGT. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Use permitted DNA substitutions to change frame A’s peptide while preserving frame B’s peptide. Report the final DNA and both frames’ original and edited peptides. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `frames` lists frame A first and frame B second. Each record supplies its inclusive DNA interval and strand. ```json { "cds": "{{cds}}", "frames": {{frames}}, "target_peptide_a": "{{target_peptide_a}}", "editable_positions": {{editable_positions}} } ``` ## Rules - Only DNA positions in `editable_positions` may change, through equal-length substitutions. - Extract the same two annotated windows before and after the edit. Reverse-complement any minus-strand window. - Use ordinary internal NCBI genetic code 11 meanings for every triplet, with no initiation override. Keep stops as `` and continue translating after them. - The final A peptide must equal `target_peptide_a`. The complete B peptide must remain exactly unchanged. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Reference material ### DNA reverse complement DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ## Answer - `cds` is the selected continuous uppercase DNA sequence. `frames` contains exactly two records, "A" then "B", in supplied window order. - `original_peptide` and `edited_peptide` are the full internal translations before and after the DNA change, retaining every stop as . `changed` is true exactly when those strings differ. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "cds": <DNA string>, "frames": [ { "frame": <"A" or "B">, "original_peptide": <amino-acid string, retaining stops as >, "edited_peptide": <amino-acid string, retaining stops as >, "changed": <boolean> } ] } ``` If no valid result exists, return: ```json { "status": "impossible", "cds": null, "frames": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Use permitted DNA substitutions to change frame A’s peptide while preserving frame B’s peptide. Report the final DNA and both frames’ original and edited peptides. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `frames` lists frame A first and frame B second. Each record supplies its inclusive DNA interval and strand. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "frames": [ { "interval": [7, 12], "strand": "+" }, { "interval": [6, 11], "strand": "+" } ], "target_peptide_a": "VI", "editable_positions": [8] } ``` ## Rules - Only DNA positions in `editable_positions` may change, through equal-length substitutions. - Extract the same two annotated windows before and after the edit. Reverse-complement any minus-strand window. - Use ordinary internal NCBI genetic code 11 meanings for every triplet, with no initiation override. Keep stops as `` and continue translating after them. - The final A peptide must equal `target_peptide_a`. The complete B peptide must remain exactly unchanged. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - `cds` is the selected continuous uppercase DNA sequence. `frames` contains exactly two records, "A" then "B", in supplied window order. - `original_peptide` and `edited_peptide` are the full internal translations before and after the DNA change, retaining every stop as . `changed` is true exactly when those strings differ. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "cds": <DNA string>, "frames": [ { "frame": <"A" or "B">, "original_peptide": <amino-acid string, retaining stops as >, "edited_peptide": <amino-acid string, retaining stops as >, "changed": <boolean> } ] } ``` If no valid result exists, return: ```json { "status": "impossible", "cds": null, "frames": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "status": "ok", "cds": "ATGAAAGTAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA", "frames": [ { "frame": "A", "original_peptide": "AI", "edited_peptide": "VI", "changed": true }, { "frame": "B", "original_peptide": "SN", "edited_peptide": "SN", "changed": false } ] }` Native CDS: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The two overlapping test windows are supplied annotations, not two natural genes. C8T changes frame A Ala→Val while preserving the overlapping frame B Ser codon AGC→AGT. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Use permitted DNA substitutions to change frame A’s peptide while preserving frame B’s peptide. Report the final DNA and both frames’ original and edited peptides. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. Coordinates and intervals are 1-based and inclusive unless a rule states otherwise. `frames` lists frame A first and frame B second. Each record supplies its inclusive DNA interval and strand. ```json { "cds": "{{cds}}", "frames": {{frames}}, "target_peptide_a": "{{target_peptide_a}}", "editable_positions": {{editable_positions}} } ``` ## Rules - Only DNA positions in `editable_positions` may change, through equal-length substitutions. - Extract the same two annotated windows before and after the edit. Reverse-complement any minus-strand window. - Use ordinary internal NCBI genetic code 11 meanings for every triplet, with no initiation override. Keep stops as `` and continue translating after them. - The final A peptide must equal `target_peptide_a`. The complete B peptide must remain exactly unchanged. ## Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. ## Answer - `cds` is the selected continuous uppercase DNA sequence. `frames` contains exactly two records, "A" then "B", in supplied window order. - `original_peptide` and `edited_peptide` are the full internal translations before and after the DNA change, retaining every stop as . `changed` is true exactly when those strings differ. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "cds": <DNA string>, "frames": [ { "frame": <"A" or "B">, "original_peptide": <amino-acid string, retaining stops as >, "edited_peptide": <amino-acid string, retaining stops as >, "changed": <boolean> } ] } ``` If no valid result exists, return: ```json { "status": "impossible", "cds": null, "frames": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `cds` is the selected continuous uppercase DNA sequence. `frames` contains exactly two records, "A" then "B", in supplied window order. `original_peptide` and `edited_peptide` are the full internal translations before and after the DNA change, retaining every stop as . `changed` is true exactly when those strings differ. Selection Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "status": "ok", "cds": <DNA string>, "frames": [ { "frame": <"A" or "B">, "original_peptide": <amino-acid string, retaining stops as >, "edited_peptide": <amino-acid string, retaining stops as >, "changed": <boolean> } ] } If no valid result exists, return: { "status": "impossible", "cds": null, "frames": null } Steps and final output Compare the designed and original DNA in both annotated reading windows. For the comparison task, represent the design as an equal-length replacement of the entire original interval: [1,n,designed_dna]. Frame A must encode target_peptide_a, and frame B must remain unchanged. Retain stop symbols () when translating these windows. Output: JSON object with exactly status, cds, frames Literal task fields { "composition_type": "pipeline.overlap_edit_verify", "output_format": "json", "scorer": "exact_match", "mode": "pipeline", "selection_rule": "Among all sequences satisfying every condition, return the lexicographically smallest full DNA string, comparing bases in the order A<C<G<T. This is a tie-break among feasible sequences, not a minimum-edit objective." } Reference material. DNA reverse complement, NCBI genetic code 11: internal DNA codons
Build and audit a controlled synonymous panel`pipeline.controlled_panel`	Generate matched synonymous CDSs with different CG targets, then report their measured RNA CG counts and DNA GC fractions.
Build and audit a controlled synonymous panel Design a panel in which each member meets its assigned coding-RNA CG count and all share the requested total number of G/C bases. Transcribe and measure each actual candidate, and compare its translated protein with the original. This connects design targets to independently checkable measurements. Practical context This prepares and audits sequences for a controlled synonymous comparison: protein identity and total GC stay fixed while CG count varies. Such panels can support experiments measuring RNA abundance or protein output. Independent measurements of each designed sequence distinguish achieved controls from merely requested ones. Method reference: Coding-sequence determinants of expression. Component tasks Build a controlled synonymous panel`design.variant_panel`Calculate total GC`stats.gc_fraction`Make the coding RNA`dna.transcribe`Count coding-RNA patterns`rna.motif_counts`Compare encoded proteins`cds.protein_equivalence` Task Outline Inputs. `cds`: `ATG AAA GCA ATT TTC GTA CTG TGA`; `target_cg_counts`: `[0, 1]`; `gc_count`: `6`. Procedure. Choose the same distinct synonymous panel as variant_panel. Recount each selected member’s overlapping RNA CG occurrences and divide its full-DNA G/C count by 24 to report the measurements. Final answer: { "status": "ok", "members": [ { "cds": "ATGAAAGCAATATTTGTACTATGA", "cg_count": 0, "gc_fraction": 0.25 }, { "cds": "ATGAAAGCAATATTCGTATTATGA", "cg_count": 1, "gc_fraction": 0.25 } ] } Checking the answer. Every sequence must satisfy its own CG target, the shared GC requirement, protein identity, and the fixed terminal-codon rule. Easy example · references supplied Prompt ## Task Create a synonymous CDS panel with the requested CG and GC counts, then report each member’s measured RNA CG count and DNA GC fraction. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "target_cg_counts": [0, 1], "gc_count": 6 } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - All returned DNA strings must be distinct. Keep members in the same order as `target_cg_counts`. - For each member, count all overlapping CG occurrences in its coding RNA, including those spanning codon boundaries. Its count must equal the corresponding requested value. - Every full DNA sequence must contain exactly `gc_count` G/C bases. - Report measurements from the final selected sequences. The GC-fraction denominator is the complete DNA length. ## Selection Among complete feasible panels, choose the lexicographically smallest list of DNA strings: compare the first differing member using A<C<G<T. Resolve choices across the whole panel, including distinctness; do not choose each member independently. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### GC fraction GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured. ## Answer - `members` follows target_cg_counts input order. Each cds is a distinct continuous uppercase synonymous DNA sequence. - `cg_count` is its integer count of overlapping CG in coding RNA. `gc_fraction` is G+C bases divided by full DNA length, as a JSON number rounded once to 6 decimal places. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "members": [ { "cds": <DNA string>, "cg_count": <nonnegative integer>, "gc_fraction": <number> } ] } ``` If no valid result exists, return: ```json { "status": "impossible", "members": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "status": "ok", "members": [ { "cds": "ATGAAAGCAATATTTGTACTATGA", "cg_count": 0, "gc_fraction": 0.25 }, { "cds": "ATGAAAGCAATATTCGTATTATGA", "cg_count": 1, "gc_fraction": 0.25 } ] }` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Create a synonymous CDS panel with the requested CG and GC counts, then report each member’s measured RNA CG count and DNA GC fraction. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "target_cg_counts": {{target_cg_counts}}, "gc_count": {{gc_count}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - All returned DNA strings must be distinct. Keep members in the same order as `target_cg_counts`. - For each member, count all overlapping CG occurrences in its coding RNA, including those spanning codon boundaries. Its count must equal the corresponding requested value. - Every full DNA sequence must contain exactly `gc_count` G/C bases. - Report measurements from the final selected sequences. The GC-fraction denominator is the complete DNA length. ## Selection Among complete feasible panels, choose the lexicographically smallest list of DNA strings: compare the first differing member using A<C<G<T. Resolve choices across the whole panel, including distinctness; do not choose each member independently. ## Reference material ### Synonymous codons Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein. ### NCBI genetic code 11: internal DNA codons Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal. ```text TTT=F TTC=F TTA=L TTG=L TCT=S TCC=S TCA=S TCG=S TAT=Y TAC=Y TAA=STOP TAG=STOP TGT=C TGC=C TGA=STOP TGG=W CTT=L CTC=L CTA=L CTG=L CCT=P CCC=P CCA=P CCG=P CAT=H CAC=H CAA=Q CAG=Q CGT=R CGC=R CGA=R CGG=R ATT=I ATC=I ATA=I ATG=M ACT=T ACC=T ACA=T ACG=T AAT=N AAC=N AAA=K AAG=K AGT=S AGC=S AGA=R AGG=R GTT=V GTC=V GTA=V GTG=V GCT=A GCC=A GCA=A GCG=A GAT=D GAC=D GAA=E GAG=E GGT=G GGC=G GGA=G GGG=G ``` ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### GC fraction GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured. ## Answer - `members` follows target_cg_counts input order. Each cds is a distinct continuous uppercase synonymous DNA sequence. - `cg_count` is its integer count of overlapping CG in coding RNA. `gc_fraction` is G+C bases divided by full DNA length, as a JSON number rounded once to 6 decimal places. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "members": [ { "cds": <DNA string>, "cg_count": <nonnegative integer>, "gc_fraction": <number> } ] } ``` If no valid result exists, return: ```json { "status": "impossible", "members": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Create a synonymous CDS panel with the requested CG and GC counts, then report each member’s measured RNA CG count and DNA GC fraction. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "ATG AAA GCA ATT TTC GTA CTG TGA", "target_cg_counts": [0, 1], "gc_count": 6 } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - All returned DNA strings must be distinct. Keep members in the same order as `target_cg_counts`. - For each member, count all overlapping CG occurrences in its coding RNA, including those spanning codon boundaries. Its count must equal the corresponding requested value. - Every full DNA sequence must contain exactly `gc_count` G/C bases. - Report measurements from the final selected sequences. The GC-fraction denominator is the complete DNA length. ## Selection Among complete feasible panels, choose the lexicographically smallest list of DNA strings: compare the first differing member using A<C<G<T. Resolve choices across the whole panel, including distinctness; do not choose each member independently. ## Answer - `members` follows target_cg_counts input order. Each cds is a distinct continuous uppercase synonymous DNA sequence. - `cg_count` is its integer count of overlapping CG in coding RNA. `gc_fraction` is G+C bases divided by full DNA length, as a JSON number rounded once to 6 decimal places. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "members": [ { "cds": <DNA string>, "cg_count": <nonnegative integer>, "gc_fraction": <number> } ] } ``` If no valid result exists, return: ```json { "status": "impossible", "members": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "status": "ok", "members": [ { "cds": "ATGAAAGCAATATTTGTACTATGA", "cg_count": 0, "gc_fraction": 0.25 }, { "cds": "ATGAAAGCAATATTCGTATTATGA", "cg_count": 1, "gc_fraction": 0.25 } ] }` Engineered trpL minigene: E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. The base minigene retains native codons 1–7 and the native terminal TGA. It is an engineered truncation, not the natural full-length gene. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Create a synonymous CDS panel with the requested CG and GC counts, then report each member’s measured RNA CG count and DNA GC fraction. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds": "{{cds}}", "target_cg_counts": {{target_cg_counts}}, "gc_count": {{gc_count}} } ``` ## Rules - Use NCBI genetic code 11 for codon interpretation. - Keep the DNA length, complete encoded protein, original first codon, and terminal stop codon unchanged. - All returned DNA strings must be distinct. Keep members in the same order as `target_cg_counts`. - For each member, count all overlapping CG occurrences in its coding RNA, including those spanning codon boundaries. Its count must equal the corresponding requested value. - Every full DNA sequence must contain exactly `gc_count` G/C bases. - Report measurements from the final selected sequences. The GC-fraction denominator is the complete DNA length. ## Selection Among complete feasible panels, choose the lexicographically smallest list of DNA strings: compare the first differing member using A<C<G<T. Resolve choices across the whole panel, including distinctness; do not choose each member independently. ## Answer - `members` follows target_cg_counts input order. Each cds is a distinct continuous uppercase synonymous DNA sequence. - `cg_count` is its integer count of overlapping CG in coding RNA. `gc_fraction` is G+C bases divided by full DNA length, as a JSON number rounded once to 6 decimal places. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "members": [ { "cds": <DNA string>, "cg_count": <nonnegative integer>, "gc_fraction": <number> } ] } ``` If no valid result exists, return: ```json { "status": "impossible", "members": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `members` follows target_cg_counts input order. Each cds is a distinct continuous uppercase synonymous DNA sequence. `cg_count` is its integer count of overlapping CG in coding RNA. `gc_fraction` is G+C bases divided by full DNA length, as a JSON number rounded once to 6 decimal places. Selection Among complete feasible panels, choose the lexicographically smallest list of DNA strings: compare the first differing member using A<C<G<T. Resolve choices across the whole panel, including distinctness; do not choose each member independently. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "status": "ok", "members": [ { "cds": <DNA string>, "cg_count": <nonnegative integer>, "gc_fraction": <number> } ] } If no valid result exists, return: { "status": "impossible", "members": null } Steps and final output For each candidate CDS, measure GC content, transcribe and count CG occurrences, and verify that translation gives the original protein. Each candidate must meet its own CG target and the shared total GC count. Include occurrences spanning codon boundaries. Output: JSON object with exactly status, members Literal task fields { "composition_type": "pipeline.controlled_panel", "output_format": "json", "scorer": "exact_match", "mode": "branch", "selection_rule": "Among complete feasible panels, choose the lexicographically smallest list of DNA strings: compare the first differing member using A<C<G<T. Resolve choices across the whole panel, including distinctness; do not choose each member independently." } Reference material. Synonymous codons, NCBI genetic code 11: internal DNA codons, Coding DNA to RNA, GC fraction
From CDS collection to a minimal measurement panel`pipeline.rna_probe_panel`	From a collection of CDSs, choose the smallest supplied RNA-probe set giving each coding RNA a distinct match signature.
From CDS collection to a minimal measurement panel Transcribe each CDS and calculate which candidate probes have exact complementary hits in each RNA. Use that match matrix to choose a minimal distinguishing panel. Report the selected probe IDs and each CDS’s binary signature in the same probe order, so the claimed discrimination can be checked directly. Practical context This joins sequence screening to an economical measurement plan: derive probe-hit signatures and choose the smallest set that distinguishes the supplied RNAs. It addresses which measurements would be informative under exact matching; binding strength and real assay performance belong to a subsequent probe-design stage. Method reference: NUPACK documentation. Component tasks Make the coding RNA`dna.transcribe`Match an RNA guide`rna.guide_hits`Distinguish CDSs with a probe panel`assay.min_probe_panel` Task Outline Inputs. `cds_panel`: `["ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ATG AAA GCC ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGT ACT TCC TGA"]`; `candidate_probes`: supplied table shown in the prompt. Procedure. Select the same minimum distinguishing probe subset as min_probe_panel. In that selected ID order, encode a 1 for an exact complementary hit and 0 otherwise for each RNA in its original input order. Final answer: { "status": "ok", "probe_ids": ["trpL_A3_window", "trpL_R12_window"], "signatures": [ { "cds_index": 1, "bits": "11" }, { "cds_index": 2, "bits": "01" }, { "cds_index": 3, "bits": "10" } ] } Checking the answer. All signatures must match sequence-derived hits, be distinct across CDSs, and come from a minimum-size panel. Easy example · references supplied Prompt ## Task Choose the smallest candidate RNA-probe panel that distinguishes every supplied CDS, and report their binary match signatures. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_panel": [ "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ATG AAA GCC ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGT ACT TCC TGA" ], "candidate_probes": { "trpL_A3_window": "GAAAAUUGCUUU", "trpL_R12_window": "AGUGCGCCACCA", "trpL_common": "ACCUUUCAGUAC" } } ``` ## Rules - Transcribe each supplied CDS and use only the candidate probes. - A probe hits an RNA if at least one equal-length window is its exact reverse complement. Use no gaps or wobble. - Record one hit/no-hit bit per selected probe, in the selected probe order. Every input CDS must have a different bit string. ## Selection Minimize the number of selected probes, then choose the lexicographically smallest sorted list of probe IDs. Compare complete ID strings in ASCII order. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNA reverse complement RNA complementary bases are A↔U and C↔G. Complement each base, then reverse the sequence to write its complementary partner 5′→3′. ## Answer - `probe_ids` contains selected IDs in ASCII order. `signatures` has one record per input CDS in input order, with its 1-based cds_index. - `bits` has one character per selected probe in probe_ids order: 1 for an exact complementary hit anywhere in that coding RNA, otherwise 0. All signature strings must be distinct. - A valid empty panel has probe_ids=[] and bits="" in every signature. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "probe_ids": [<probe ID string>], "signatures": [ { "cds_index": <positive integer>, "bits": <string of 0 and 1> } ] } ``` If no valid result exists, return: ```json { "status": "impossible", "probe_ids": null, "signatures": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "status": "ok", "probe_ids": ["trpL_A3_window", "trpL_R12_window"], "signatures": [ { "cds_index": 1, "bits": "11" }, { "cds_index": 2, "bits": "01" }, { "cds_index": 3, "bits": "10" } ] }` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Panel order: native trpL, synonymous Ala3 GCA→GCC variant, synonymous Arg12 CGC→CGT variant. Candidate 12-base probes are designed reverse complements of native windows 4–15, 28–39, and 16–27. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Choose the smallest candidate RNA-probe panel that distinguishes every supplied CDS, and report their binary match signatures. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_panel": {{cds_panel}}, "candidate_probes": {{candidate_probes}} } ``` ## Rules - Transcribe each supplied CDS and use only the candidate probes. - A probe hits an RNA if at least one equal-length window is its exact reverse complement. Use no gaps or wobble. - Record one hit/no-hit bit per selected probe, in the selected probe order. Every input CDS must have a different bit string. ## Selection Minimize the number of selected probes, then choose the lexicographically smallest sorted list of probe IDs. Compare complete ID strings in ASCII order. ## Reference material ### Coding DNA to RNA The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA. ### RNA reverse complement RNA complementary bases are A↔U and C↔G. Complement each base, then reverse the sequence to write its complementary partner 5′→3′. ## Answer - `probe_ids` contains selected IDs in ASCII order. `signatures` has one record per input CDS in input order, with its 1-based cds_index. - `bits` has one character per selected probe in probe_ids order: 1 for an exact complementary hit anywhere in that coding RNA, otherwise 0. All signature strings must be distinct. - A valid empty panel has probe_ids=[] and bits="" in every signature. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "probe_ids": [<probe ID string>], "signatures": [ { "cds_index": <positive integer>, "bits": <string of 0 and 1> } ] } ``` If no valid result exists, return: ```json { "status": "impossible", "probe_ids": null, "signatures": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Hard example · reference knowledge assumed Prompt ## Task Choose the smallest candidate RNA-probe panel that distinguishes every supplied CDS, and report their binary match signatures. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_panel": [ "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ATG AAA GCC ATT TTC GTA CTG AAA GGT TGG TGG CGC ACT TCC TGA", "ATG AAA GCA ATT TTC GTA CTG AAA GGT TGG TGG CGT ACT TCC TGA" ], "candidate_probes": { "trpL_A3_window": "GAAAAUUGCUUU", "trpL_R12_window": "AGUGCGCCACCA", "trpL_common": "ACCUUUCAGUAC" } } ``` ## Rules - Transcribe each supplied CDS and use only the candidate probes. - A probe hits an RNA if at least one equal-length window is its exact reverse complement. Use no gaps or wobble. - Record one hit/no-hit bit per selected probe, in the selected probe order. Every input CDS must have a different bit string. ## Selection Minimize the number of selected probes, then choose the lexicographically smallest sorted list of probe IDs. Compare complete ID strings in ASCII order. ## Answer - `probe_ids` contains selected IDs in ASCII order. `signatures` has one record per input CDS in input order, with its 1-based cds_index. - `bits` has one character per selected probe in probe_ids order: 1 for an exact complementary hit anywhere in that coding RNA, otherwise 0. All signature strings must be distinct. - A valid empty panel has probe_ids=[] and bits="" in every signature. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "probe_ids": [<probe ID string>], "signatures": [ { "cds_index": <positive integer>, "bits": <string of 0 and 1> } ] } ``` If no valid result exists, return: ```json { "status": "impossible", "probe_ids": null, "signatures": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Expected answer `Final answer: { "status": "ok", "probe_ids": ["trpL_A3_window", "trpL_R12_window"], "signatures": [ { "cds_index": 1, "bits": "11" }, { "cds_index": 2, "bits": "01" }, { "cds_index": 3, "bits": "10" } ] }` Engineered trpL derivative(s): E. coli K-12 MG1655 trpL, locus `b1265`. Reference CDS: `NC_000913.3`, bases 1,323,038–1,323,082 on the minus strand. Protein: `NP_415781.1`. Panel order: native trpL, synonymous Ala3 GCA→GCC variant, synonymous Arg12 CGC→CGT variant. Candidate 12-base probes are designed reverse complements of native windows 4–15, 28–39, and 16–27. Shared source record and original sequence. Retrieved 26 October 2026. Template · unfilled placeholders Template ## Task Choose the smallest candidate RNA-probe panel that distinguishes every supplied CDS, and report their binary match signatures. ## Inputs DNA inputs are 5′→3′ coding strands. Ignore whitespace between bases. ```json { "cds_panel": {{cds_panel}}, "candidate_probes": {{candidate_probes}} } ``` ## Rules - Transcribe each supplied CDS and use only the candidate probes. - A probe hits an RNA if at least one equal-length window is its exact reverse complement. Use no gaps or wobble. - Record one hit/no-hit bit per selected probe, in the selected probe order. Every input CDS must have a different bit string. ## Selection Minimize the number of selected probes, then choose the lexicographically smallest sorted list of probe IDs. Compare complete ID strings in ASCII order. ## Answer - `probe_ids` contains selected IDs in ASCII order. `signatures` has one record per input CDS in input order, with its 1-based cds_index. - `bits` has one character per selected probe in probe_ids order: 1 for an exact complementary hit anywhere in that coding RNA, otherwise 0. All signature strings must be distinct. - A valid empty panel has probe_ids=[] and bits="" in every signature. Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. ### Output structure ```text { "status": "ok", "probe_ids": [<probe ID string>], "signatures": [ { "cds_index": <positive integer>, "bits": <string of 0 and 1> } ] } ``` If no valid result exists, return: ```json { "status": "impossible", "probe_ids": null, "signatures": null } ``` After any explanation, write `Final answer:` on its own line. Begin the raw JSON on the next line; it may span multiple lines and ends your response. Answer format and selection · json `output_format = json` `probe_ids` contains selected IDs in ASCII order. `signatures` has one record per input CDS in input order, with its 1-based cds_index. `bits` has one character per selected probe in probe_ids order: 1 for an exact complementary hit anywhere in that coding RNA, otherwise 0. All signature strings must be distinct. A valid empty panel has probe_ids=[] and bits="" in every signature. Selection Minimize the number of selected probes, then choose the lexicographically smallest sorted list of probe IDs. Compare complete ID strings in ASCII order. Output structure Use exactly this structure. `<…>` marks a typed value; array items and map entries repeat as needed. { "status": "ok", "probe_ids": [<probe ID string>], "signatures": [ { "cds_index": <positive integer>, "bits": <string of 0 and 1> } ] } If no valid result exists, return: { "status": "impossible", "probe_ids": null, "signatures": null } Steps and final output Transcribe each CDS and check every probe against every resulting RNA using guide_hits with max_mismatches=0. Record a binary match table, then choose the smallest set of probes that gives each RNA a different match pattern. Break ties using sorted probe IDs. Output: JSON object with exactly status, probe_ids, signatures Literal task fields { "composition_type": "pipeline.rna_probe_panel", "output_format": "json", "scorer": "exact_match", "mode": "pipeline", "selection_rule": "Minimize the number of selected probes, then choose the lexicographically smallest sorted list of probe IDs. Compare complete ID strings in ASCII order." } Reference material. Coding DNA to RNA, RNA reverse complement

Reference material

These definitions and lookup tables supply the biological information used by easy prompts. Each task links only to the entries it needs. Instance data and required benchmark conventions are given in both prompt variants.

DNA reverse complement

DNA complementary bases are A↔T and C↔G. Complement each base, then reverse the sequence to write the opposite strand 5′→3′.

Coding DNA to RNA

The supplied DNA is the coding strand, written 5′→3′. Copy its bases in the same order, replacing each T with U to obtain the coding RNA.

Codon length

A codon is a group of three consecutive nucleotides.

Residue-to-codon correspondence

Each amino-acid residue is encoded by one consecutive codon. The initiating residue is residue 1; a stop codon encodes no residue.

NCBI genetic code 11: internal DNA codons

Each internal DNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal.

TTT=F  TTC=F  TTA=L  TTG=L
TCT=S  TCC=S  TCA=S  TCG=S
TAT=Y  TAC=Y  TAA=STOP  TAG=STOP
TGT=C  TGC=C  TGA=STOP  TGG=W
CTT=L  CTC=L  CTA=L  CTG=L
CCT=P  CCC=P  CCA=P  CCG=P
CAT=H  CAC=H  CAA=Q  CAG=Q
CGT=R  CGC=R  CGA=R  CGG=R
ATT=I  ATC=I  ATA=I  ATG=M
ACT=T  ACC=T  ACA=T  ACG=T
AAT=N  AAC=N  AAA=K  AAG=K
AGT=S  AGC=S  AGA=R  AGG=R
GTT=V  GTC=V  GTA=V  GTG=V
GCT=A  GCC=A  GCA=A  GCG=A
GAT=D  GAC=D  GAA=E  GAG=E
GGT=G  GGC=G  GGA=G  GGG=G

Annotated CDS initiation

The annotated first codon of an intact CDS encodes initiating methionine (M), even when its internal-codon assignment is different.

NCBI genetic code 11: allowed DNA starts

The allowed annotated initiation codons are TTG, CTG, ATT, ATC, ATA, ATG, and GTG.

NCBI genetic code 11: DNA stop codons

The DNA stop codons are TAA, TAG, and TGA. All other unambiguous DNA codons are sense codons.

NCBI genetic code 11: internal RNA codons

Each internal RNA codon is an aligned group of three bases. The table gives its one-letter amino-acid assignment; STOP marks a stop signal.

UUU=F  UUC=F  UUA=L  UUG=L
UCU=S  UCC=S  UCA=S  UCG=S
UAU=Y  UAC=Y  UAA=STOP  UAG=STOP
UGU=C  UGC=C  UGA=STOP  UGG=W
CUU=L  CUC=L  CUA=L  CUG=L
CCU=P  CCC=P  CCA=P  CCG=P
CAU=H  CAC=H  CAA=Q  CAG=Q
CGU=R  CGC=R  CGA=R  CGG=R
AUU=I  AUC=I  AUA=I  AUG=M
ACU=T  ACC=T  ACA=T  ACG=T
AAU=N  AAC=N  AAA=K  AAG=K
AGU=S  AGC=S  AGA=R  AGG=R
GUU=V  GUC=V  GUA=V  GUG=V
GCU=A  GCC=A  GCA=A  GCG=A
GAU=D  GAC=D  GAA=E  GAG=E
GGU=G  GGC=G  GGA=G  GGG=G

Initiation after DNA editing

At the first DNA codon, TTG, CTG, ATT, ATC, ATA, ATG, and GTG encode initiating M under NCBI genetic code 11. If the first edited codon is not one of these initiators, use its ordinary internal assignment.

Initiation of coding RNA

At the first RNA codon, UUG, CUG, AUU, AUC, AUA, AUG, and GUG encode initiating M under NCBI genetic code 11. If that first codon is not an allowed initiator, use its ordinary internal assignment.

IUPAC DNA base sets

An IUPAC symbol permits any base in its listed set. Choices at different positions are independent.

A=A  C=C  G=G  T=T
R=AG  Y=CT  S=CG  W=AT  K=GT  M=AC
B=CGT  D=AGT  H=ACT  V=ACG  N=ACGT

Synonymous codons

Synonymous codons encode the same amino acid. Replacing a codon with a synonymous codon preserves that residue in the protein.

GC fraction

GC fraction is the number of G and C bases divided by the number of bases in the sequence being measured.

Codon adaptation index

For n included codons with supplied positive weights w₁,…,wₙ, CAI = (w₁ × … × wₙ)^(1/n), the geometric mean of those weights.

Canonical RNA base pairs

The canonical Watson–Crick RNA pairs are A–U, U–A, C–G, and G–C.

RNA reverse complement

RNA complementary bases are A↔U and C↔G. Complement each base, then reverse the sequence to write its complementary partner 5′→3′.

Dot-bracket notation

Matching parentheses mark paired RNA positions; a dot marks an unpaired position. Each closing parenthesis pairs with the most recent unmatched opening parenthesis.

Trypsin cleavage specificity

Trypsin cleaves after K or R, except when the next residue is P. The K or R remains on the upstream peptide.

RNase T1 cleavage specificity

RNase T1 cleaves RNA after G. That G remains on the upstream fragment.

N-glycosylation sequon

The conventional N-glycosylation sequon is N–X–S or N–X–T, where X is any amino-acid residue except P.

Monoisotopic isotope masses

Use these isotope masses in Da as exact decimal constants. For a neutral, unmodified linear peptide, sum neutral free-amino-acid molecular formulas and subtract one H2O per peptide bond. Equivalently, sum incorporated residue masses and add one H2O per complete peptide. Do not round intermediate values.

12C = 12.00000000000
1H = 1.00782503223
14N = 14.00307400443
16O = 15.99491461957
32S = 31.9720711744

Monoisotopic residue masses

Use these incorporated residue masses in Da as exact decimal constants. Each is the corresponding neutral free amino acid minus one H2O. A free linear peptide has the sum of its residue masses plus one H2O, 18.01056468403 Da. Round only the final peptide mass.

A=71.03711378515  C=103.00918495955  D=115.02694302429  E=129.04259308875
F=147.06841391407  G=57.02146372069  H=137.05891185847  I=113.08406397853
K=128.09496301519  L=113.08406397853  M=131.04048508847  N=114.04292744138
P=97.05276384961  Q=128.05857750584  R=156.10111102405  S=87.03202840472
T=101.04767846918  V=99.06841391407  W=186.07931295073  Y=163.06332853364

Source: NIST isotope masses, snapshot accessed 2 November 2026. The incorporated residue masses are derived from the neutral free-amino-acid molecular formulas minus H2O, using the listed central isotope values without rounding. Measurement uncertainties are not propagated; the displayed decimal values are exact benchmark constants. Source and retrieval notes are documentation and are not included in model prompts.

Kyte–Doolittle hydropathy scores

Use these Kyte–Doolittle residue scores. Whole-protein mean hydropathy is their arithmetic mean, counting every residue occurrence.

A=1.8  R=-4.5  N=-3.5  D=-3.5
C=2.5  Q=-3.5  E=-3.5  G=-0.4
H=-3.2  I=4.5  L=3.8  K=-3.9
M=1.9  F=2.8  P=-1.6  S=-0.8
T=-0.7  W=-0.9  Y=-1.3  V=4.2

Ionization and net charge

A basic site contributes +1/(1+10^(pH−pKa)); an acidic site contributes −1/(1+10^(pKa−pH)).

Basic sites: the N terminus and K, R, H side chains. Acidic sites: the C terminus and D, E, C, Y side chains. Count each terminus once and each listed side chain once per residue occurrence; sum their contributions.

Sources and practical precedents

NCBI genetic code 11 — Codon meanings and annotated initiators.
INSDC feature-table specification — CDS annotations and nucleotide ambiguity conventions.
ExPASy PeptideCutter — Protease specificity models and their experimental limits.
DNA Chisel examples — Translation-preserving motif removal, local GC constraints, and codon optimization.
DNA Chisel specifications — Explicit sequence constraints, including codon choices and translation preservation.
LinearDesign — Joint optimization of codon adaptation and a thermodynamic RNA-folding objective.
ViennaRNA — RNA structure analysis using thermodynamic models; the benchmark’s pair-count and additive-score exercises use simpler rules.
Primer3 — Primer selection with melting-temperature, complementarity, and pair-quality criteria.
GoldenMutagenesis domestication — Silent removal of internal cleavage sites for Golden Gate assembly.
Twist gene-design resources — Synthesis complexity associated with sequence repeats, homopolymers, and GC extremes.
Codon Statistics Database — Comparative codon-usage statistics, including GC and third-position GC.
ExPASy ProtParam documentation — Protein composition, hydropathy, theoretical pI, and other sequence descriptors.
ExPASy PeptideMass documentation — Theoretical proteolytic digests and peptide masses for protein identification.
NIST Atomic Weights and Isotopic Compositions — Source of the fixed 12C, 1H, 14N, 16O, and 32S central masses used to derive the residue and water constants; snapshot accessed 2 November 2026. The mass reference entry records the derivation and numerical convention.
PROSITE N-glycosylation pattern — The sequon pattern and the distinction between a candidate site and actual modification.
Cytiva ion-exchanger selection guide — Protein charge, pI, operating pH, and ion-exchange resin selection.
RNA and mRNA sequence mapping by mass spectrometry — RNase digestion, fragment identification, and complementary mapping information for RNA characterization.
NUPACK documentation — Thermodynamic analysis and design of nucleic-acid structures and interacting complexes.
CHARMING — Harmonizing synonymous codon usage to reproduce a desired usage pattern.
Restriction-site markers in engineered alleles — An example of adding a silent restriction site to help identify a desired genome edit.
Origin and evolution of the zinc finger antiviral protein — A biological example of RNA recognition sensitive to CpG dinucleotides.
RNAi-resistant rescue in a B-cyclin study — Synonymous changes used to make rescue cDNAs resistant to the corresponding RNAi reagents.
Coding-sequence determinants of expression — A synonymous GFP library illustrates experimentally varying coding RNA while preserving the protein; it is not the benchmark’s exact CG/GC-controlled panel.
Site-directed mutagenesis protocol — Point mutations, deletions, and other explicitly designed sequence changes.
Addgene bacterial expression collection — Examples of protein-expression constructs with fusion and purification tags.
Removal of affinity tags with TEV protease — TEV-mediated tag removal and the resulting protein products.
SwiftLib — Degenerate-codon-library optimization for desired amino-acid coverage and manageable library size.
RNase T1: enzyme specificity and applications — Thermo Fisher’s enzyme documentation: guanosine-specific cleavage of single-stranded RNA; fungal origin and RNA-analysis uses.
EcoRI recognition and cleavage site — A real restriction enzyme recognizing GAATTC; cleavage is G/AATTC.
Kyte–Doolittle hydropathy scale — The original experimentally informed scale for displaying hydrophilic and hydrophobic regions of proteins.
Structural basis for microRNA targeting — Argonaute uses guide RNA base pairing to recognize target mRNA; a uniform mismatch count is a simpler benchmark rule.
Nussinov and colleagues: Algorithms for Loop Matchings — The 1978 dynamic-programming treatment of noncrossing base-pair matching, a historical foundation for simplified RNA-folding problems.
ApoB RNA editing: CAA becomes a stop codon — Experimental C-to-U RNA editing changes a glutamine CAA codon to UAA in apolipoprotein B mRNA.
The apoB-editing catalytic subunit is a cytidine deaminase — Identification of the enzyme chemistry underlying apoB RNA editing.
Programmable editing by endogenous ADAR — Primary research on ADAR-catalyzed adenosine-to-inosine editing. The benchmark’s translation-only example represents the recoding as A→G.
Synonymous mutations and restoration of mRNA base pairing — The rpsT study found that compensatory synonymous substitutions restoring predicted pairing could partly or fully restore fitness in some tested mutants.
NCBI CDKN2A annotation — p16INK4a and p14ARF transcripts share a second exon read in different frames, providing a concrete biological example of overlapping coding information.
Imaging individual mRNA molecules with labeled probes — Primary research on detecting mRNA through complementary oligonucleotide probes; assay chemistry and fluorescence are outside the benchmark.
His6-tagged protein purification — A laboratory teaching resource explaining binding of a six-histidine tag to nickel affinity material.
Ensembl calculated variant consequences — Rule-based annotation using codons and amino acids, with synonymous, missense, stop-gained, and frameshift consequence categories.

Evaluation design

Paired prompts, separate score axes

Each underlying instance gets easy and hard prompts with one shared answer and validator. Track reference support separately from sequence length, number of edits, constraint density, or search-space size. Keep tool access and token budgets matched.

Report canonical final-answer accuracy and answer-format compliance separately. A correct answer must match the normalized reference, including every requested field. Intermediate results may also receive diagnostic scores; a plausible explanation cannot rescue a wrong final answer.

Prevent sequence leakage

Split related sequence families before deriving tasks. Keep parent CDSs, homologs, recodings, mutants, RNA products, protein products, and composed examples in the same partition.

Hold out selected task combinations as a separate composition-generalization test. Do not use the exact intermediate sequences from a held-out composition to train its individual tasks.

Certify reference answers

Generate one canonical reference answer per instance, including every tie-break. Independently verify protein preservation, all constraints, and the required selection rule. Keep exact-search instances small enough to certify the chosen feasible or optimal answer, not just one witness.

A reference may say IMPOSSIBLE only after an independent solver certifies infeasibility. NONE means no qualifying readout exists. A search timeout establishes neither result.

Status of this proposal

This document defines candidate task types, prompt templates, and scoring rules. It does not claim that a dataset, reference implementation, solver, or benchmark score already exists.

Before release: generate instances, implement reference solvers and independent validators, verify canonical-answer uniqueness and easy/hard equivalence, and pilot task lengths and search limits. The illustrative examples have been checked; they are not a released benchmark dataset.

CDS-bench: Task definitions

Six examples from the 84 tasks

Translate DNA to protein

Disrupt a specified local hairpin

Digest with trypsin rules

Remove specified DNA motifs

From CDS panel to a diagnostic peptide

From specified RNA edits to protein

What an instance can supply

What the benchmark does not ask for

Easy and hard variants

Reference supplied

Reference knowledge assumed

Task schema and scoring

Evaluating a response

Individual tasks (65)

CDS interpretation

Reverse-complement a CDS

Role in related tasks

Task Outline

Make the coding RNA

Role in related tasks

Task Outline

Translate DNA to protein

Task Outline

Validate a complete CDS

Task Outline

Map a nucleotide to its codon

Role in related tasks

Task Outline

Output structure

Map residues back to DNA

Role in related tasks

Task Outline

Sequence variation

Apply a DNA edit script

Role in related tasks

Task Outline

Classify a substitution effect

Practical context

Task Outline

Determine frame preservation

Practical context

Task Outline

Output structure

Compare encoded proteins

Practical context

Task Outline

Resolve an ambiguous codon

Role in related tasks

Task Outline

Assess an ambiguous stop

Role in related tasks

Task Outline

Sequence statistics

Count sense codons

Role in related tasks

Task Outline

Output structure

Calculate total GC

Practical context

Task Outline

Calculate third-position GC

Practical context

Task Outline

Locate GC-window violations

Practical context

Task Outline

Calculate codon adaptation index

Task Outline

Find nucleotide patterns

Role in related tasks

Task Outline

Output structure

Protein analysis

Find a supplied peptide pattern

Role in related tasks

Task Outline

Locate N-glycosylation sequons

Practical context