itsvenu / signatures-manuscript Goto Github PK

Dear Venu,
I hope this message finds you well. I really appreciated your great manuscript about mutational signatures in pediatric cancer, since I am also working on MSs in pediatric leukemia.
I read that you used a DKFZ in-house pipeline for variant calling, but also combined results with Mutect2. I am using Mutect2 (tumor-normal mode) for variant calling and I have encountered many issues with tumor-in-normal contamination and found active regions being completely excluded from Mutect2 somatic calling (normal-control samples were contaminated in some cases).
Of course, this is very deleterious for the subsequent MSs detection, since many true somatic variants are not called by Mutect2 and missed in the signature matrixes generation.
Did you also faced this issue? Is your pipeline optimized in order to perform a better calling of somatic variants, using matched controls and maybe whitelisting some known hotspots?
Thanks in advance for your kind answer.
Alberto