GATK18.104.22.168 Mutect2 TumorOnly mode miss variant observed in BAM file.
I'm using GATK Mutect2 to perform variant calling using tumor sample. Here is something strange that I cannot understand.
When I called variant using a full target region BED file, one SNP within a particular target region is missed even though 19% bases on that location are mutated as shown in the following IGV screenshot.
And when I called variant using just that location ( -L chr3:179,234,277-179,234,317), this variant is reported by Mutect2 as expected:
chr3 179234297 . A G . . DP=293;ECNT=1;POP_AF=5.000e-08;TLOD=146.06 GT:AD:AF:F1R2:F2R1:MBQ:MFRL:MMQ:MPOS:SA_MAP_AF:SA_POST_PROB 0/1:221,63:0.230:111,32:110,31:27:170,179:60:20:0.202,0.212,0.222:7.973e-03,0.013,0.979
I wonder why specifying a larger target region would cause miss of this obvious SNV and what would you do to prevent this kind of loss?