HGVD: Human Genetic Variation Database update 2026 with sequencing of 4344 genetically diverse Japanese individuals

Population-scale genome resources remain biased toward European ancestry, limiting variant interpretation in underrepresented populations. Here, we expanded the Human Genetic Variation Database (HGVD), a free Japanese allele-frequency resource launched in 2013. HGVD v3.1 integrates 1209 exomes with whole-genome sequencing (WGS) data from 3135 Japanese individuals selected by principal component analysis (PCA) to span genetic diversity, providing aggregate variant statistics for 4344 individuals, 3248 genotyped individuals, and expression quantitative trait loci (eQTL) from 300 individuals. For research and clinical use, HGVD provides coordinated JBrowse-based browsers for GRCh37/hg19, GRCh38/hg38, and T2T-CHM13. The non-GRCh37 variant tracks were generated by liftOver from the GRCh37 call set. Variant frequencies, sample counts, and coverage are available across assemblies, whereas eQTL annotations are currently available only on hg19. A downloadable, graph-genome-ready structural-variant release is provided, and the new pathSTR portal reports population-level distributions for 68 medically relevant tandem-repeat loci, derived by applying STRdust to long-read Oxford Nanopore data from 1019 individuals of the 1000 Genomes Project remapped to T2T-CHM13. HGVD v3.1 supports population-matched interpretation of rare pathogenic and pharmacogenomic variants, comparison of repeat-expansion alleles with long-read reference distributions, and structural-variant-aware reference construction, supporting precision medicine for Japan and East Asian populations. HGVD is available at https://www.hgvd.genome.med.kyoto-u.ac.jp/.