I am performing a variant calling analysis and I have some genes that present a high amount of SNPs and INDELs, such as some HLA genes. But, actually, this regions present a lot of alterations, so I would like to discard them, because I think they might be increasing the noise of my results. The problem is that I don't know which regions are hypermutated and which of them actually present a high frequency of variants in my cohort but are not normally hypermutated.
How can I deal with this ? Thanks!