This track shows small variants (SNVs and short indels) that were observed in affected or case individuals of disease-study cohorts, annotated with their predicted protein consequence and colored by severity. It is one half of a matched pair: the companion Population reference track shows the same kind of variants seen in population reference cohorts and in unaffected relatives or controls. Displaying the two together lets you compare, for example, how often a loss-of-function variant in a gene of interest is seen in affected individuals versus the general/unaffected background. For the full list of contributing projects, see the SNV Frequencies collection page.
The affected counts are drawn from the affected or case arm of five disease-study cohorts: SFARI SPARK WES and SFARI SPARK WGS (autism spectrum disorder probands), SCHEMA (schizophrenia cases), GREGoR (affected rare-disease participants), and GA4K (a pediatric -rare-disease cohort). For SPARK, SFARI WGS, SCHEMA, and GREGoR the source data carries an -explicit affected/unaffected (or case/control) label and only the affected arm feeds this +rare-disease cohort). For SPARK, SFARI WGS, SCHEMA, and GREGoR, the source data carries an +explicit affected/unaffected (or case/control) label, and only the affected arm feeds this track. GA4K reports a single cohort-wide frequency with no per-individual label; because it -is a rare-disease cohort it is counted as affected here, with the caveat that it enrolls +is a rare-disease cohort, it is counted as affected here, with the caveat that it enrolls parent-child trios, so a minority of its carriers are unaffected parents. Genotyping-array cohorts are not included in either track.
Variants are colored by their most severe predicted consequence:
| Color | Consequence class | Examples |
|---|---|---|
| Protein-truncating / loss-of-function | stop_gained, frameshift, splice_donor, splice_acceptor, stop_lost, start_lost | |
| Missense / in-frame | missense, inframe_insertion, inframe_deletion, protein_altering | |
| Synonymous | synonymous, stop_retained | |
| Non-coding / intergenic | intron, non_coding, intergenic, UTR |
The score (used for shading) is the pooled affected/case allele frequency times 1000.
Affected AF is the pooled rate across contributing affected arms:
affectedAF = sum(AC) / sum(AN), where affectedAC sums the allele counts
-and affectedAN sums the allele numbers across each cohort/arm that ships both AC and
+and affectedAN sums the allele numbers across each cohort/arm that provides both AC and
AF (the per-arm AN is derived as round(AC / AF)). Cohorts that publish only AF
contribute via a configured default_an in the build configuration. Cohorts
that publish only AC and have no default_an set (currently GREGoR's per-arm
AC_AFFECTED/UNAFFECTED/UNKNOWN) are listed in affectedCohorts but do not contribute
to the pool numerator or denominator; their carriers are visible in the per-database AC
column instead. The pooled rate is preferred over a max-across-cohorts statistic so a
small cohort with a high local AF cannot dominate the displayed frequency.
To look for protein-truncating variants that are common in affected individuals but rare in the background, set the Consequence filter to Stop Gained, Frameshift, Splice Donor and Splice Acceptor (these appear red), then add an upper limit on the Background AF filter. Each variant here carries both its affected frequency and its background frequency, so this isolates variants seen in cases with little or no presence in the population/unaffected set. Comparing visually against the Population reference track shows the same contrast across a whole gene.
Variant-frequency VCFs from the contributing cohorts were stripped of unneeded INFO fields,
normalized with bcftools norm (splitting multi-allelic sites), and merged with
bcftools merge. The merged callset was annotated with predicted protein
consequences using bcftools csq against the
Ensembl
GRCh38 release 115 gene models.
A custom Python script (vcfToBigBed.py) then read the per-cohort allele
counts and frequencies and, for each variant, pooled the allele counts and allele numbers
across the affected arms (case/proband subgroups, plus GA4K whole-cohort) to produce this
track, and across the population cohorts and unaffected/control subgroups to produce the
companion Population reference track. A variant
seen in both groups appears in both tracks. The build is documented in the
makeDoc, and the scripts are on
GitHub.
Because the merged callset combines cohorts whose redistribution licenses differ, this track is not available for download and is not in the Table Browser. It can be reconstructed from the individual source VCFs using the conversion scripts and the build documentation. The per-project subtracks on the SNV Frequencies collection page document how to obtain each source dataset.
This track is only possible thanks to the data from the participants and families of the SFARI SPARK, SCHEMA, GREGoR and GA4K studies. Click the individual project subtracks on the SNV Frequencies collection page for the specific credits and citations of each cohort. Thanks to Alex Ioannidis, UCSC, for the inspiration for this track and to Andreas Lahner, MGZ, for feedback.
For the primary citation of each source cohort, see the References section on the SNV Frequencies collection page. The merged-track build uses the following tools:
Danecek P, McCarthy SA. BCFtools/csq: haplotype-aware variant consequences. Bioinformatics. 2017 Jul 1;33(13):2037-2039. PMID: 28205675; PMC: PMC5870570