detecting LoH in non-cancer samples

Question

[clementine]

Hello everyone,

I was wondering if someone can suggest a way to detect loss of heterozygosity from the wes data. I've found a number of tools but they're mostly designed for tumor samples which is not my case. I have wes data of patients and am interested to see if we have haploinsufficiency of some known loci due to deletion of one copy of the gene.

I was thinking maybe first detecting ROH and if they overlap with my loci, then checking for coverage around that region. But I'm not sure if when we have only one copy of the gene, it would definitely alter the coverage in wes.

Any suggestions?

Thanks in advance!

Answer 1

[bernatgel]

If you expect a deletion of 1 on the copies I'd go first with a copy-number calling (there are good tools working on WES) and then check if the variant frequencies (lack of heterozygous calls) support the call (you can use CNVfilteR for that). I think it'll be much easier and sensitive than the other way around.