Mutect2 somatic hemizygote variant on chrX

Question

[cocchi.e89]

I am working on calling some somatic variants from WES samples, I followed step-by-step the GATK guidelines. This is not a tumor sample so I don't compare the exome with another but with a multisaples panel-of-normals. Basically at the end of all filterings etc. you got a VCF file with PASS on likely somatic variants.

I got this result in an hemizygotic variant on chromosome X X-107865086-G-A:

X   107865086   .   G   A   .   PASS    CONTQ=93;DP=31;ECNT=1;GERMQ=65;MBQ=0,34;MFRL=0,166;MMQ=60,60;MPOS=19;POPAF=7.30;ROQ=93;SEQQ=93;STRANDQ=93;TLOD=99.53    GT:AD:AF:DP:F1R2:F2R1:SB    0/1:0,31:0.970:31:0,15:0,16:0,0,18,13

I noticed that this variant was called by GATK Haplotypecaller too, and is an hemizygosis on a male patient in chrX. How is that possible that this variant is considered somatic by all the Mutect2 pipeline?? I don't think a VAF of 97% can ever be somatic....

Thank you very much in advance for any help!

[igor]

I don't think a VAF of 97% can ever be somatic

Maybe not likely, but certainly possible. In this case, if this is a male, then you only have one chrX.

[cocchi.e89]

Yes sorry Igor, I meant not really likely