This site is now read-only. You can find our new documentation site and support forum for posting questions here.
Be sure to read our welcome blog!
Adopter Trimming Vs Non Trimming in Mutect2
Dear gatk Team
We have performed mutation calling (using Mutect2-gatk 18.104.22.168) on samples where 1) adopter trimming was NOT performed and, 2) adopter trimming was performed. We have observed some major differences in calls with or without adopter trimming data.
Gene Name Chr cDNA change Amino acid change cfDNA frequency Total cfDNA reads Germline frequency Mutation Type Cosmic ID SV-GENOME-POSITION
PMS2 7 c.C287T p.(Ala96Val) 1.4 445 0.2223 SNV chr7:6043387:G>A
AKT1 14 c.A191T p.(Lys64Met) 1.5 293 0.5278 SNV chr14:105243092:T>A
UNC13D 17 c.G2034T p.(Lys678Asn) 1.9 243 0.6255 SNV chr17:73830756:C>A
MN1 22 c.G788A p.(Arg263His) 2.6 248 0.6375 SNV chr22:28195744:C>T
With Trimming **
When we do perform adopter trimming as compared to Non trimming, we mostly do lose calls. For further investigation, bamout files for both the scenarios were checked and to our surprise the bamout file with trimmed adopter doesn't have the active region at all. I have attached the image where 6 bams files are loaded in IGV. 3 for NON trimmed and 3 for trimmed data. It can be seen that the mutation lies in the middle region of read (and not toward the ends where soft clipped bases are). In both the situations only 2 C-->A mutations (**gene UNC13D **in above table) are present but active region is missing in trimmed data.
Its very confusing that what makes mutect2 to think that its an active region when there are adopters and what makes it not an active region when there are no adopters.
Can I get clarity on this for this please?