This track shows allele counts and frequencies from gnomAD-Canada, the Canadian arm of the Genome Aggregation Database. The v1.0 release aggregates whole-genome sequencing data from 10,487 participants in the Canadian COVID-19 HostSeq databank, a national initiative launched in April 2020 that recruited people infected with SARS-CoV-2 together with controls from fourteen clinical and epidemiological studies across Canada. Because the cohort was assembled to study host susceptibility to a common infection rather than a rare genetic disease, it serves here as a population-scale reference of Canadian genetic variation.
Frequencies are reported for the whole cohort and for nine genetic-ancestry groups inferred by the gnomAD pipeline: African/African-American, Latino/Admixed-American, European (non-Finnish), Ashkenazi Jewish, East Asian, European (Finnish), Middle Eastern, South Asian, and Remaining individuals. The release contains about 213 million variant sites. The per-variant details page also shows the group with the highest frequency (grpmax) and the gnomAD variant-quality filter status.
This is a Variant Call Format (VCF) track. Hovering over a variant shows the overall allele frequency, and clicking a variant opens a details page with the overall and per-ancestry allele counts, allele numbers and frequencies, the dbSNP identifier, and the variant filter status. The filter status is one of PASS, RF (failed the gnomAD random-forest variant-quality filter), or InbreedingCoeff (excess heterozygosity); variants with zero release allele count (AC0) are not shown.
The data can be explored interactively with the Table Browser or the Data Integrator. For programmatic access, our REST API can be used; the track name is hostseq. For bulk download, the VCF is available from our download server. The original Hail table can be downloaded from the gnomAD-Canada portal.
Whole-genome sequencing was performed on 10,487 HostSeq participants at three CGEn sites on the Illumina NovaSeq 6000 platform at 30X coverage. Reads were aligned to GRCh38 with the DRAGEN mapper (DRAGMAP v1.3.0), de-duplicated with Picard (v2.25.0) and base-quality-recalibrated with GATK (v4.2.5.0). Per-sample SNVs and indels were called with GATK HaplotypeCaller in DRAGEN mode, and per-sample gVCFs were joint-genotyped with GATK GenotypeGVCFs. The joint callset was then run through the Broad Institute's gnomAD QC v3 pipeline (adapted for on-premises use at the BC Cancer Genome Sciences Centre), which performs sample quality control, relatedness and genetic-ancestry inference by principal-component analysis, random-forest variant filtering, and computes overall and per-ancestry allele frequencies. Full pipeline details are in the release README and the HostSeq cohort reference below.
The release-sites Hail table (release_sites_hostseq.tar.gz, ~101 GB) was downloaded from https://www.bcgsc.ca/gnomad/download/release_sites_hostseq.tar.gz and read with Hail. The overall and per-ancestry allele counts, allele numbers and frequencies were flattened from the table's freq array into VCF INFO fields, together with the dbSNP rsID and the variant filter status, and exported as a sites-only VCF. The source table is already biallelic-split and uses UCSC chr-prefixed contigs, so no allele-splitting or chromosome renaming was needed. Of the 232,028,491 variant rows in the release table, 213,382,086 with a non-zero release allele count were written; the 18,646,405 AC0 sites (no observed allele in the high-quality release) were dropped. The source covers the autosomes plus chrX and chrY. About 15% of the written variants (32,074,697) carry the gnomAD RF quality-filter flag and 158,230 the InbreedingCoeff flag; the remaining 181,149,159 are PASS. Filter-flagged variants are retained so the full release is browsable, and can be hidden with the track's filter controls. Processing documentation is in the varFreqs makeDoc file; the conversion script is in our GitHub scripts directory.
Data were generated by the gnomAD-Canada team as part of the Canadian Genomic Data Commons (CGDC) and the Pan-Canadian Genome Library (PCGL), using CGEn's HostSeq databank (funded by the Government of Canada through Genome Canada, Project DACO-18). Thanks to the HostSeq participants and to the gnomAD-Canada project leadership (Jordan Lerner-Ellis, Steven Jones and Daniel Taliun) for making the aggregate data publicly available. Per the data providers, no attempts should be made to relabel the gnomAD-Canada genetic-ancestry groups.
Yoo S, Garg E, Elliott LT, Hung RJ, Halevy AR, Brooks JD, Bull SB, Gagnon F, Greenwood C, Lawless JF et al. HostSeq: a Canadian whole genome sequencing and clinical data resource. BMC Genom Data. 2023 May 2;24(1):26. PMID: 37131148; PMC: PMC10152008
Chen S, Francioli LC, Goodrich JK, Collins RL, Kanai M, Wang Q, Alföldi J, Watts NA, Vittal C, Gauthier LD et al. A genomic mutational constraint map using variation in 76,156 human genomes. Nature. 2024 Jan;625(7993):92-100. PMID: 38057664; PMC: PMC11629659