Agent skill

bio-blast-searches

Run remote BLAST searches against NCBI databases using Biopython Bio.Blast. Use when identifying unknown sequences, finding homologs, or searching for sequence similarity against NCBI's nr/nt databases.

View SKILL.md on GitHub Repository
Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/data/blast-searches

SKILL.md

BLAST Searches

Run BLAST searches against NCBI databases using Biopython's Bio.Blast module.

Required Import

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

BLAST Programs

Program Query Database Use Case
blastn Nucleotide Nucleotide DNA/RNA sequence similarity
blastp Protein Protein Protein sequence similarity
blastx Nucleotide Protein Find protein hits for DNA query
tblastn Protein Nucleotide Find DNA encoding protein-like
tblastx Nucleotide Nucleotide Translated vs translated

Core Function

NCBIWWW.qblast()

Submit a BLAST query to NCBI servers.

python
from Bio.Blast import NCBIWWW

# Simple BLASTN search
result_handle = NCBIWWW.qblast('blastn', 'nt', sequence)

Key Parameters:

Parameter Description Example
program BLAST program 'blastn', 'blastp'
database Target database 'nr', 'nt', 'refseq_rna'
sequence Query sequence String or SeqRecord
entrez_query Limit by Entrez query 'Homo sapiens[organism]'
hitlist_size Max hits to return 50
expect E-value threshold 0.001
word_size Word size 11 for blastn
gapcosts Gap penalties '5 2' (open, extend)
format_type Output format 'XML' (default), 'Text'

Common Databases

Nucleotide:

Database Description
nt All GenBank + EMBL + DDBJ
refseq_rna RefSeq RNA sequences
refseq_genomic RefSeq genomic sequences

Protein:

Database Description
nr Non-redundant protein
refseq_protein RefSeq proteins
swissprot SwissProt (curated)
pdb Protein structures

Parsing Results

NCBIXML Parser

python
from Bio.Blast import NCBIWWW, NCBIXML

# Run BLAST
result_handle = NCBIWWW.qblast('blastn', 'nt', sequence)

# Parse XML results
blast_record = NCBIXML.read(result_handle)
result_handle.close()

# Iterate hits
for alignment in blast_record.alignments:
    print(f"Hit: {alignment.title}")
    for hsp in alignment.hsps:
        print(f"  E-value: {hsp.expect}")
        print(f"  Score: {hsp.score}")
        print(f"  Identity: {hsp.identities}/{hsp.align_length}")

Alignment/HSP Attributes

python
# Alignment (hit) attributes
alignment.title          # Hit description
alignment.accession      # Accession number
alignment.length         # Subject sequence length
alignment.hsps           # List of HSPs

# HSP (High-scoring Segment Pair) attributes
hsp.score               # Raw score
hsp.bits                # Bit score
hsp.expect              # E-value
hsp.identities          # Number of identical positions
hsp.positives           # Number of positive-scoring positions
hsp.gaps                # Number of gaps
hsp.align_length        # Alignment length
hsp.query               # Aligned query sequence
hsp.match               # Match line (| for identity)
hsp.sbjct               # Aligned subject sequence
hsp.query_start         # Query start position
hsp.query_end           # Query end position
hsp.sbjct_start         # Subject start position
hsp.sbjct_end           # Subject end position
hsp.strand              # Strand (blastn)
hsp.frame               # Reading frame (blastx/tblastn)

Code Patterns

Basic BLASTN

python
from Bio.Blast import NCBIWWW, NCBIXML

sequence = '''ATGAAAGCAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA'''

print("Running BLASTN (this may take a minute)...")
result_handle = NCBIWWW.qblast('blastn', 'nt', sequence)

blast_record = NCBIXML.read(result_handle)
result_handle.close()

print(f"\nFound {len(blast_record.alignments)} hits")
for alignment in blast_record.alignments[:5]:
    hsp = alignment.hsps[0]
    print(f"\n{alignment.title[:70]}...")
    print(f"  E-value: {hsp.expect:.2e}")
    print(f"  Identity: {hsp.identities}/{hsp.align_length} ({100*hsp.identities/hsp.align_length:.1f}%)")

BLASTP with Organism Filter

python
from Bio.Blast import NCBIWWW, NCBIXML

protein_seq = '''MVLSPADKTNVKAAWGKVGAHAGEYGAEALERMFLSFPTTKTYFPHFDLSH'''

result_handle = NCBIWWW.qblast(
    'blastp',
    'nr',
    protein_seq,
    entrez_query='Mammalia[organism]',
    hitlist_size=20,
    expect=0.001
)

blast_record = NCBIXML.read(result_handle)
result_handle.close()

for alignment in blast_record.alignments[:10]:
    hsp = alignment.hsps[0]
    print(f"{alignment.accession}: E={hsp.expect:.2e} - {alignment.title[:50]}...")

BLAST from FASTA File

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

record = SeqIO.read('query.fasta', 'fasta')

result_handle = NCBIWWW.qblast('blastn', 'nt', record.seq)
blast_record = NCBIXML.read(result_handle)
result_handle.close()

for alignment in blast_record.alignments[:5]:
    print(f"{alignment.accession}: {alignment.title[:60]}...")

Save Results to File

python
from Bio.Blast import NCBIWWW

result_handle = NCBIWWW.qblast('blastn', 'nt', sequence)

# Save XML for later parsing
with open('blast_results.xml', 'w') as out:
    out.write(result_handle.read())
result_handle.close()

# Parse saved file
from Bio.Blast import NCBIXML
with open('blast_results.xml') as f:
    blast_record = NCBIXML.read(f)

Extract Top Hits

python
def get_top_hits(sequence, program='blastn', database='nt', num_hits=10, evalue=0.01):
    result_handle = NCBIWWW.qblast(
        program, database, sequence,
        hitlist_size=num_hits,
        expect=evalue
    )
    blast_record = NCBIXML.read(result_handle)
    result_handle.close()

    hits = []
    for alignment in blast_record.alignments:
        hsp = alignment.hsps[0]
        hits.append({
            'accession': alignment.accession,
            'title': alignment.title,
            'evalue': hsp.expect,
            'identity': hsp.identities / hsp.align_length,
            'coverage': hsp.align_length / blast_record.query_length
        })
    return hits

hits = get_top_hits(my_sequence, num_hits=5)
for hit in hits:
    print(f"{hit['accession']}: {hit['identity']:.1%} identity, E={hit['evalue']:.2e}")

BLASTX (DNA to Protein)

python
from Bio.Blast import NCBIWWW, NCBIXML

dna_sequence = '''ATGAAAGCAATTTTCGTACTGAAAGGTTGGTGGCGCACTTCCTGA'''

result_handle = NCBIWWW.qblast('blastx', 'nr', dna_sequence)
blast_record = NCBIXML.read(result_handle)
result_handle.close()

for alignment in blast_record.alignments[:5]:
    hsp = alignment.hsps[0]
    print(f"{alignment.accession}: frame {hsp.frame}, E={hsp.expect:.2e}")
    print(f"  {alignment.title[:60]}...")

Multiple Sequences

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

def blast_multiple(fasta_file, program='blastn', database='nt'):
    results = {}
    for record in SeqIO.parse(fasta_file, 'fasta'):
        print(f"BLASTing {record.id}...")

        result_handle = NCBIWWW.qblast(program, database, str(record.seq), hitlist_size=5)
        blast_record = NCBIXML.read(result_handle)
        result_handle.close()

        results[record.id] = blast_record
        time.sleep(5)  # Be nice to NCBI servers

    return results

Filter by Identity/Coverage

python
def filter_blast_hits(blast_record, min_identity=0.9, min_coverage=0.8):
    query_length = blast_record.query_length
    filtered = []

    for alignment in blast_record.alignments:
        for hsp in alignment.hsps:
            identity = hsp.identities / hsp.align_length
            coverage = hsp.align_length / query_length

            if identity >= min_identity and coverage >= min_coverage:
                filtered.append({
                    'accession': alignment.accession,
                    'title': alignment.title,
                    'identity': identity,
                    'coverage': coverage,
                    'evalue': hsp.expect
                })

    return filtered

Common Errors

Error Cause Solution
Timeout NCBI servers busy Retry later, use smaller query
No hits Wrong database or program Check program/database match
Empty results E-value too stringent Increase expect parameter
URLError Network issue Check connection, retry

Timing Notes

  • Remote BLAST can take 30 seconds to several minutes
  • Longer queries take longer
  • Peak times (US business hours) may be slower
  • For many queries, consider local BLAST

Decision Tree

Need to BLAST a sequence?
├── DNA query?
│   ├── Against DNA database? → blastn
│   └── Against protein database? → blastx
├── Protein query?
│   ├── Against protein database? → blastp
│   └── Against DNA database? → tblastn
├── Want to limit organisms?
│   └── Use entrez_query parameter
├── Need many searches?
│   └── Consider local-blast skill
└── Need results fast?
    └── Use local-blast skill

Related Skills

  • local-blast - Run BLAST locally for faster, unlimited searches
  • entrez-fetch - Fetch full records for BLAST hits
  • sequence-io - Read query sequences from files

Recommended Agent Skills

Expand your agent's capabilities with these related and highly-rated skills.

majiayu000/claude-skill-registry

agent-ops-spec

Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-state

Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-spec

Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-testing

Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-testing

Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.

163 31
Explore
majiayu000/claude-skill-registry

agent-ops-state

Maintain .agent state files. Use at session start, after meaningful steps, and before concluding: read/update constitution/memory/focus/issues/baseline consistently.

163 31
Explore

Didn't find tool you were looking for?

Be as detailed as possible for better results