My ReadPosRankSumTest (ReadPosRankSum) distribution looks funny

Hi,

I was looking at the hard filtering cutoffs and hence looking at the distributions of MQ, QD, MRankSum and so on. All look fine - similar to yours (https://software.broadinstitute.org/gatk/documentation/article.php?id=6925) but one. I was wondering wether my ReadPosRankSumTest (ReadPosRankSum) distribution looks indeed funny or if there is a specific reason for that.
Thanks!

Tagged:

Best Answer

Answers

  • JulsJuls Member
    edited April 2018

    Hi Sheila,
    Thank you for you answer. It's standard Illumina data. HiSeq 2000. What kind of "quirk" could it be? Any suggestions as to the cause? So you think I could just ignore it? I figured I use a more stringent cutoff e.g, -2.5.

  • SheilaSheila Broad InstituteMember, Broadie, Moderator admin

    @Juls
    Hi,

    I don't want to say you can ignore it for sure.

    Have you done any QC on your data? For example, have you checked the coverage, contamination, chimerism, and sequencing artifact metrics?

    Is this human data? Have you done some concordance tests against truth sets?

    Thanks,
    Sheila

Sign In or Register to comment.