bio-grep 0.1.5

Local-first biological sequence search engine

bio-grep

bio-grep is a local-first biological search engine for fast iterative querying over FASTA, FASTQ, GFF3, GTF, BED, VCF, SAM/BAM, GenBank, and EMBL corpora.

Install

pip install "bio-grep[bam]"

or

uv tool install "bio-grep[bam]"

One-line installer (latest):

curl -fsSL https://raw.githubusercontent.com/ravishar313/bio-grep/main/scripts/install.sh | bash

Optional version pin:

curl -fsSL https://raw.githubusercontent.com/ravishar313/bio-grep/main/scripts/install.sh | bash -s -- --version 0.1.5

Quick Start

Index one or more files:

bio-grep index data/fasta/*.fasta --index-dir . --k 7

Run a motif search:

bio-grep search --motif "C..C" --top-k 20 --index-dir .

Search annotation formats:

bio-grep search --file-type gff --region "chr1:120-300" --feature-type-exact gene --top-k 20 --index-dir .
bio-grep search --file-type vcf --region "chr1:1-1000000" --filter-exact PASS --top-k 20 --index-dir .
bio-grep search --file-type sam --seqid-exact chr1 --mapq-gte 30 --top-k 20 --index-dir .

Fetch full sequence payload by ID:

bio-grep fetch s1 --index-dir .

Highlights

• Agent-first compact jsonl output by default.
• Stable metadata keys (null when unavailable).
• Rich filter support for sequence and annotation search.
• --version flag for CLI/version pinning workflows.

Python API

from pathlib import Path
from bio_grep.engine import BioGrep

bio = BioGrep(base_dir=Path("."), k=7)
bio.index([Path("data/fasta/uniprot_reviewed_human_250.fasta")])
rows = bio.search(motif="C..C", length_lt=200).top_k(10)
print(rows[0]["id"], rows[0]["score"])

Project Links

• Repository: https://github.com/ravishar313/bio-grep
• Issues: https://github.com/ravishar313/bio-grep/issues