This track collection contains structural variant (SV) and copy-number variant (CNV) callsets derived from Illumina short-read sequencing. Most SV tracks in the browser now come from long-read platforms (see the companion Long-read SVs supertrack); the short-read callsets here are included as comparators so users can evaluate the extra sensitivity of long-read calls and cross-check a variant across technologies.
SV length statistics (min / median / max) are computed from the svLen field of each track, in base pairs. For the Abel CCDG callset, a large fraction of records are breakend (BND) translocations where svLen=-1 is used as a sentinel, which shows up in both min and median.
| Dataset | N samples | Cohort / disease | Sequencing | SVs | Min | Median | Max |
|---|---|---|---|---|---|---|---|
| CCDG 17,795 | 17,795 | NHGRI CCDG + PAGE + SGDP (B38 native + B37 lifted) | Illumina short-read (LUMPY + CNVnator + svtyper) | 737,998 | -1 | -1 | 217,985,413 |
| 1KG 3202 | 3,202 | 1000 Genomes expanded cohort | Illumina short-read (GATK-SV) | 173,366 | 1 | 314 | 154,807,729 |
| ToMMo 48K CNV | 48,874 | Japanese, general population | Illumina short-read (GATK CNV, 1 kb bins, shown as two bigWigs) | ~2M bins with CNV carriers; not comparable to per-SV counts above | |||
Site-frequency callset from 17,795 deeply sequenced genomes (Abel et al. 2020, Nature; PMID 32460305). Two non-overlapping public releases are combined in this track: the B38 callset (14,623 samples called natively on GRCh38) and the B37 callset (8,417 samples, lifted). Variants are colored by SV type (DEL / DUP / INV / MEI / BND) and carry per-population allele counts for eight ancestry groups plus a HIGH/LOW confidence filter.
1000 Genomes 3202-sample Illumina short-read GATK-SV callset (Byrska-Bishop et al. 2022). 173,366 SVs across 7 classes (DEL, INS, DUP, INV, CPX, CNV, CTX) with AC/AN/AF and per-superpopulation AFs (AFR/AMR/ASN/EUR/SAN).
Per-1 kb-bin copy-number carrier counts from short-read whole-genome sequencing of 48,874 Japanese individuals (jMorp 48KJPN-CNV Frequency Panel, release 20230828), called with GATK CNV germline workflows. Shown as a multiWig overlay: red = samples with copy-number loss (CN<2) per bin, green = samples with gain (CN>2) per bin. This is a useful short-read point of comparison to the ToMMo 333-sample long-read SV track under the Long-read SVs supertrack.
See the Data Access section of each subtrack's page for download links. Build documentation lives alongside the scripts at doc/hg38/srSv.txt; conversion scripts and autoSql schemas are at makeDb/scripts/srSv.
Each subtrack credits its respective upstream project; see the individual description pages.
See the individual subtrack description pages for the specific references.