New SNV Frequencies container track on hg38
Luis Nassar
We are pleased to announce a new SNV Frequencies container track on the human assembly (GRCh38/hg38). The collection unifies single-nucleotide variant allele frequencies from 30+ population resequencing and biobank projects worldwide, covering roughly 1.7 million genomes, exomes, and genotyping arrays, into a single place where a variant's frequency can be compared across populations, ancestries, and disease cohorts.
The collection includes three combined tracks that aggregate the source data along different lines:
- Population reference (default view): 1.24 billion variants pooled across general-population and biobank cohorts plus the unaffected/control arms of the disease studies.
- Disease cohorts: 133 million variants in the affected or case arms of five disease-study cohorts (SFARI SPARK WES and WGS, SCHEMA, GREGoR, GA4K). Each variant also carries its background frequency for easy case-vs-population comparison.
- Genotyping Array Databases Combined: 14.7 million variants from three array cohorts (TPMI Taiwan, Mexico Biobank, UK Biobank imputed), kept separate from the sequencing-based tracks because chip data has different per-variant confidence.
rs4986893, a CYP2C19 stop-gained variant common in East Asian populations. The mouseover ranks ToMMo Japan, KOVA Korea, and WBBC China at the top, well above the 1.6% pooled background AF.
Each variant in the combined tracks is colored by predicted protein consequence (loss-of-function, missense, synonymous, non-coding), and carries pooled allele counts and pooled allele frequencies (sum AC / sum AN) across the contributing cohort arms. Cross-database filters allow you to narrow to variant type, consequence, source database, length, or any combination of per-track AC/AF/AN ranges.
The container track pulls together cohorts from across the world. A high-level summary of the regions and contributing projects is shown below; a complete table with per-cohort sample counts, data types, sub-populations and download status is on the container track description page.
| Region | Contributing cohorts |
|---|---|
| East Asia | ToMMo, GenomeAsia, NPM, KOVA, WBBC, ChinaMAP, TPMI |
| South Asia | IndiGen, GenomeIndia |
| Africa | Tishkoff |
| Americas | AllOfUs, TOPMed, ABraOM, Mexico Biobank |
| Europe | FinnGen, SweGen, GoNL, HRC, UK Biobank, ALFA |
| Middle East | Saudi |
| Oceania | MGRB |
| Disease cohorts | SFARI SPARK (WES + WGS), SCHEMA, GREGoR, GA4K |
| Global reference panels | SGDP, gnomAD HGDP+1kG |
| Long-read / linked-read | GA4K, CoLoRSdb, SVatalog |
License restrictions on some sources limit redistribution; see the container track description page for per-cohort details.
We plan to continue updating this track as more population-scale allele-frequency datasets become available. If you are involved with a project that publishes variant frequencies and would like to contribute, please reach out.
We would like to thank the millions of participants worldwide who donated samples and shared health information to make this data possible; the investigators and data-access teams who provided the source variant files and helped shape the track, including Matthew Hansen and Sarah Tishkoff (Tishkoff lab, Penn) for the Tishkoff180 African cohort, Insu Jang (KOBIC) for KOVA, Yanan Cao (Shanghai) for ChinaMAP, Matthew Hobbs (Garvan) for MGRB, Adam Ameur and Johan Viklund for SweGen, Ameena Suliman and Julia Sommer (SFARI) for SPARK data access, and Cole Shanks and Qudsi Aljabiri (Ioannidis lab, UCSC) for the All of Us callset; Andreas Lahner (MGZ) for feedback on the design; and Alex Ioannidis (UCSC) for the motivation behind the track.