BESDQ — Fast BESD eQTL Query Tool

Fast queries of BESD (Binary Efficient Sequential Data) eQTL summary statistics. Supports direct file access and an optimised SQLite index for repeated queries.

Why BESDQ?

• No database server required — queries run directly against BESD files or a local SQLite index
• SMR-compatible output — drop-in replacement for SMR query workflows
• 10–200× faster than direct BESD access when using the SQLite index for repeated queries
• Compact storage — index is 50–70% the size of the original BESD files

See docs/performance-westra.md for benchmarks.

Installation

git clone https://github.com/explodecomputer/besdq.git
cd besdq
pip install -e .

Quickstart

Query BESD files directly (no setup):

besdq --beqtl-summary data/westra_eqtl_hg19 \
  --snp-chrpos 1:100000-2000000 \
  --probe-chrpos 1:1000000-2000000 \
  --out results/output

Query via SQLite index (faster for repeated queries):

# One-time index creation
besdq --beqtl-summary data/westra_eqtl_hg19 --index data/westra_eqtl_hg19.db

# Query the index
besdq --besd-index data/westra_eqtl_hg19.db \
  --snp-chrpos 1:100000-2000000 \
  --probe-chrpos 1:1000000-2000000 \
  --out results/output

Full usage reference: docs/usage.md

Documentation

Audience	Document
Analysts	docs/usage.md — installation, all query modes, Python API
Developers	docs/developer.md — architecture, file format, design decisions
Evaluators	docs/performance.md — benchmarks

Credits

Built on the BESD format from the SMR software by Jian Yang's group.

License

MIT