OpenSkillIndex← back to search
claude-code

BioClaw blast-search

Run BLAST sequence similarity searches. Use when the user asks to BLAST a sequence, find similar sequences, identify a gene/protein, or do homology search. Triggers on "blast", "sequence similarity", "homology", "identify sequence".

install
source · Clone the upstream repo
git clone https://github.com/Runchuan-BU/BioClaw
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/Runchuan-BU/BioClaw "$T" && mkdir -p ~/.claude/skills && cp -r "$T/container/skills/blast-search" ~/.claude/skills/runchuan-bu-bioclaw-blast-search && rm -rf "$T"
manifest: container/skills/blast-search/SKILL.md
source content

BLAST Search

Run NCBI BLAST+ searches inside the BioClaw container.

When to Use

  • User provides a DNA/RNA/protein sequence and wants to find similar sequences
  • User asks to identify an unknown sequence
  • User wants to check sequence conservation across species

How to Execute

1. Determine BLAST program

InputDatabaseProgram
Nucleotide queryNucleotide DB
blastn
Protein queryProtein DB
blastp
Nucleotide queryProtein DB
blastx
Protein queryNucleotide DB
tblastn

2. For local BLAST (sequences provided by user)

# Create query file
cat > /tmp/query.fa << 'EOF'
>query_sequence
ATGCGATCGATCGATCG...
EOF

# Create subject file (if user provides reference)
cat > /tmp/subject.fa << 'EOF'
>reference
ATGCGATCGATCGATCG...
EOF

# Run BLAST
blastn -query /tmp/query.fa -subject /tmp/subject.fa -outfmt 6 -evalue 1e-5

3. For remote BLAST (against NCBI databases)

Use BioPython's NCBIWWW module:

from Bio.Blast import NCBIWWW, NCBIXML
from Bio import SeqIO

# Read sequence
sequence = "ATGCGATCGATCGATCG..."

# Run remote BLAST
result_handle = NCBIWWW.qblast("blastn", "nt", sequence)
blast_records = NCBIXML.parse(result_handle)

for record in blast_records:
    for alignment in record.alignments[:10]:
        print(f"Title: {alignment.title}")
        for hsp in alignment.hsps:
            print(f"  Score: {hsp.score}, E-value: {hsp.expect}")
            print(f"  Identity: {hsp.identities}/{hsp.align_length} ({hsp.identities/hsp.align_length*100:.1f}%)")

4. Output format

Present results in a clear table:

*BLAST Results (top 10 hits)*

• Hit 1: Homo sapiens TP53 gene (98.5% identity, E=1e-45)
• Hit 2: Mus musculus Trp53 gene (89.2% identity, E=1e-38)
...

5. Follow-up suggestions

After showing results, suggest:

  • Multiple sequence alignment of top hits
  • Phylogenetic analysis
  • Domain/motif analysis of the query
  • Structural comparison if protein