possible contamination

In verbose output mode, MuTect gives reasons for rejected calls, and one of them is "possible contamination".
Does anyone know if it is a contamination of the tumoral sample by the normal sample, or a contamination with another patient sample?
I couldn't find documentation about that.


  • kcibulkcibul Cambridge, MAMember, Broadie, Dev ✭✭✭

    This is contamination with another patient sample. The default cross-patient contamination level for MuTect is 2%, but can be adjusted (ideally informed by a method like ContEst) to a more accurate setting.

  • ltononltonon Member

    Thanks I understand a bit better.
    But this leads to another question. I ran ContEst on my data and the average contamination given was 1.5. How can I have mutations rejected by MuTect if its cross-patient contamination level by default is 2%?
    Can you explain me how MuTect decides to reject a mutation because of "possible contamination" ?
    Thanks a lot

Sign In or Register to comment.