Data downloads
The datasets behind the site, as JSON. Built entirely on public sources (GenomeIndia, 1000 Genomes, PGS Catalog, gnomAD, AlphaMissense) — see the methods for how each was computed and its limitations.
The pooled all-India allele frequencies come from GenomeIndia (10,000 Indian whole genomes, IBDC, GRCh38). We publish only the small set of derived per-variant numbers we extract, with a GenomeIndia citation — we do not re-host the raw summary file. GenomeIndia's open release is pooled all-India only; the per-subpopulation view remains 1000 Genomes (smaller samples).
Reuse is welcome; a formal open-data license (CC BY 4.0) and a versioned/DOI release are being set up. Until then, please cite as below (and cite GenomeIndia for the all-India frequencies), and email if you plan to build on it.
Effect-allele frequencies for curated variants: a pooled All-India column (IN) from GenomeIndia (~10,000 Indian genomes, GRCh38), alongside the 5 South-Asian 1000G subpopulations (PJL/GIH/BEB/ITU/STU) and SAS/EUR/EAS/AFR/AMR super-populations for finer resolution.
Per-PGS × per-population mean-shift (SD) and mis-stratification % (share above the European top-10% cutoff) for 6 polygenic scores, including a pooled All-India (GenomeIndia) row.
The two-tier research-gap score over all 1,648 diseases: burden × South-Asian mis-calibration × tractability, with every component score and the per-lead why/study/learn.
AlphaMissense-pathogenic missense variants observed in South Asians, flagged for European-absent and ClinVar-uncertain reclassification candidates.
AlphaMissense / SIFT / PolyPhen predictions and SAS-vs-EUR enrichment for the curated variant panel.
1,049 medical terms with plain + student-level definitions — the accessibility layer, reusable on its own.
How to cite
D2I2: Decoding Disease in India. India Whitespace Score and genomic transferability atlas. d2i2.org (accessed 2026-07-12). Pooled all-India allele frequencies derived from GenomeIndia (Nat Genet 2025).