gatk
gatk copied to clipboard
broadinstitute
mutect2 homozygous alt GT annotation
Feature request
Running mutect2 (GATK v4.2.2) in tumor only mode on WES samples and asking it to use/keep germline and PON mutations. I have noticed that mutect never sets the GT field to be 1/1 (homozygous alt). even with 0 ref and 300 alt reads it will say 0/1, is there a way to set a threshold to define a point on which not to call heterozygosity?
Thanks!
In a recent tumor-only analysis I also observed this. I could not find any documentation why no hom_alt will be reported. There is some dicussion in Chapter II E in "Notes on MUTECT2" but it is still puzzling to me. I assume it is related to the fact that homozygous calls - as opposed to heterozygous calls - do not follow a "traceable" 50% ratio and could therefore also be of somatic origin.
Yes I read this document but it still doesn't make sense to me either..
i was thinking of using a tool such as bcftools +setGT to set the right GT myself with some kind of tule like: bcftools +setGT $FILE -- -t q -n 1 -i'AD[*:1]=0 & AF>0.93' & GT='0:1' But it seems it can set every value except homozygosity.. -n 1 here.
@jkobject Discrete genotypes are not meaningful in somatic calling and Mutect2 has long simply emitted 0/1 for all variants as a placeholder. We're probably going to get rid of the GT entirely soon.
Hello @davidbenjamin ,
Why do you say they aren't meaningful? The phasing of a somatic mutations seems quite important to me.
What about when I ask mutect2 to emit germlines and pon sites? Aren't you doing phasing?
Best,