Biopython repo skill

Use this skill when a task asks for Biopython, Bio.*, BioSQL, or general computational-biology workflows that Biopython owns: sequences, biological file formats, alignments, BLAST/search output parsing, phylogenetic trees, macromolecular structures, public biological databases, motifs, restriction enzymes, and specialized bioinformatics modules.

First checks

• Public install: python -m pip install biopython; conda users can install from conda-forge.
• Source/developer install: python -m pip install -e . from a Biopython checkout when editing the package.
• Required runtime dependency: NumPy. Optional features may need ReportLab, matplotlib, networkx, rdflib, database drivers, or external bioinformatics executables.
• Minimal import check:

import Bio
from Bio.Seq import Seq
from Bio import SeqIO
print(Bio.__version__, Seq("ATGGCC").translate())

• Run scripts/biopython_quick_smoke.py when you need a safe offline package smoke check before deeper work.
• Read references/repo-provenance.md before deciding whether this skill is current for a checkout. If the commit, package version, or evidence paths differ substantially, refresh the skill.
• Read references/troubleshooting.md for install/import/build issues, optional dependency failures, online-service constraints, and source-tree warning recovery.
• Read references/capability-map.md when a request spans multiple Biopython modules or you need optional dependency ownership.

Route by task

User task	Read next	Notes
Create/manipulate Seq, MutableSeq, SeqRecord, annotations, features, locations, codon tables, reverse complements, translations, GC/protein utilities	sequence-objects-and-features	In-memory object semantics; route file parsing/writing onward to file I/O.
Parse, read, write, index, or convert FASTA/FASTQ/GenBank/EMBL/SwissProt/UniProt XML/alignment files	file-io-and-format-conversion	Covers SeqIO, AlignIO, low-level FASTA/FASTQ iterators, BGZF, index, index_db, and format names.
Pairwise/multiple alignments, substitution matrices, BLAST/SearchIO parsing, local or online BLAST result handling, phylogenetic tree I/O/traversal	alignment-search-and-phylogeny	For online BLAST policy, also read the web/database sub-skill.
PDB/mmCIF/BinaryCIF/PQR/PDBML structure parsing, SMCRA traversal, atom/residue selection, disordered atoms, contacts, geometry, superposition	structural-bioinformatics	External tools such as DSSP/NACCESS/MSMS are optional and not part of the base install.
Entrez, KEGG, UniProt, Swiss-Prot, ExPASy, GenBank, Medline, GEO, public-database parsers, qblast network etiquette, BioSQL	web-databases-and-biosql	Default to offline parsing examples; live services require user email/API policy and network handling.
Motifs, PWM/PSSM, JASPAR, restriction enzymes, clustering, phenotype arrays, GenePop/popgen, GenomeDiagram/graphics, SeqUtils/ProtParam, long-tail modules	specialized-analyses-and-graphics	Optional graphics/database integrations are documented but not required for base use.

Boundary rules

• Do not use Biopython as a replacement for HTSlib-backed BAM/CRAM/VCF command wrappers; use dedicated HTS tools when the request needs samtools/bcftools/tabix semantics.
• Do not claim online-service verification unless you actually ran the network call with user-approved email/API key/rate policy.
• Do not require the original Biopython repository checkout for runtime guidance; this skill bundles the operating references and smoke scripts future agents need.
• If a workflow needs an optional package, external executable, database server, or credentials, stop at the sub-skill troubleshooting section and ask for/verify that dependency before running live work.