Revisiting NF2-related variants with chromosome maps
This image details all clinvar variants in chromosomes of interest (1, 2, 5, 7, 10, 11, 14-17, 19, 20, 22, X), based on genes of interest. variants in nf2-related genes highlighted á la opacity heatmap in purple.
Recently, I participated with some coworkers in a genomics hackathon hosted by SVAI and the NF2 Project.
Our team explored data from Clinvar (what is clinvar? why should i trust its content?) and tried to see if we could coax out patterns while checking if these same variants in genes associated with the NF2 condition were present in Onno’s (the subject’s) DNA sequence.
I decided to revisit some of the work we did using a nice dataset extracted from Clinvar by mapping out these variants, because the dataset didn’t get fully used (an hg19-based VCF filtered down so that it contains variants in genes related to the NF2 gene, based on BioGrid interactions). I was interested in one column — clinical significance.
Variant Clinical Significance
0 - Uncertain significance, 1 - not provided, 2 - Benign, 3 - Likely benign, 4 - Likely pathogenic, 5 - Pathogenic, 6 - drug response, 7 - histocompatibility, 255 - other
This can be treated as a discrete measure of how pathogenic a variant is.
In these chromosome maps, 255 is considered a “0″. “7″ was not actually found.
Each slice represents the 14 chromosomes that contain NF2-associated genes with their variants mapped by chromosome coordinate, and their global location mapped by a Clinvar pathogenicity label (at one end, only ~200 variants in total were tagged with a CLNSIG of 6, which is why it is sparser than the others).
Visually, certain angles make these bins look like groupings of stars.
Since we’re highlighting variants from entire genes linked to the NF2 gene, we expect to see clusters of purple start to gather. It’s interesting to explore how those clusters separate by the labelled pathogenicity of those variants.