search

SACGF / variantgrid

VariantGrid public repo
Other
23 stars 2 forks source link

Suggestion: Improving 'Nearby Variants' Page #641

Closed a1618617 closed 2 years ago

a1618617 commented 2 years ago

Hi guys,

I'm doing some re-classifications for the GAS cohort paper, and I quite liked Varsome's way of summarising Clinvar variants (Benign-Pathogenic) across "hotspots" (domains for regions with uniprot ID, or +/-25bp flanking regions for un-annotated protein regions):

Example: https://varsome.com/variant/hg19/6%3A31747470%20G%3EA? image

It'll be really cool to add that information to the variant page/nearby variants page. Example, on the variant page you have the number of clinvar variants within the exon/domain/50bp flanking, but in brackets if you could put the number of (n_Benign, n_LB, n_VUS, n_LP, n_P), that would be really useful for med scientists to quickly assess the PM1 criteria: image

Thanks.

davmlaw commented 2 years ago

I'd already made this change but forgot to actually use it by passing clinical_significance=True to the function (d'oh!).

I've turned it on and also enabled it in SA Path

a1618617 commented 2 years ago

Thanks Dave!

davmlaw commented 2 years ago

Noticed a minor bug, even if there are no classifications it shows: Classifications: (),

davmlaw commented 2 years ago

Had to undo change in SA Path as it had diverged a bit from prod, so didn't work right

davmlaw commented 2 years ago

Tested vg test

http://test.variantgrid.com/variantopedia/view_variant/959143

Shows: ClinVar: (B: 1, LB: 1),