DuckDB build of the complete BRaVa gene-level association table: 61,791,444 rows covering every gene x trait x variant-mask x MAF-cutoff x ancestry cell, with the Burden, SKAT and SKAT-O p-values, the IVW Burden effect size and its standard error, and the cross-cohort heterogeneity test.
Built by etl/build_db.py from the 280 published phenotype/{P}.{ANC}.json files (the same data as the 19,541 per-gene files, in the pivot that costs two orders of magnitude fewer requests upstream). Sorted for run-length encoding, no ART indexes: 0.87 GB, and every query in the README stays under 70 ms.
Corresponds to the upstream data release of 12 Aug 2026. Summary statistics only, not for clinical use. Source data: https://github.com/nikbaya/brava_browser (MIT).