Speed improvement of MuTect2


We are using MuTect2 for calling of somatic variants in RNAseq. So we only want to look at coding regions. I tried the -L flag for only coding regions but it was not improving the speed of MuTect2. Is this because of the mathematics behind it? Is there another way to improve the speed?

Thank you


  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie admin
    Hi @Hecate, unfortunately MuTect 2 is very slow in its present form. We are working on speeding it up, but in the meantime one thing you can do is to run it in parallel over the intervals of interest, i.e. use what we call scatter gather parallelism.
Sign In or Register to comment.