VQSR Training

atksatks Posts: 16Member
edited October 2012 in Ask the GATK team

When performing VQSR, the data set has its variants overlapped with the training set, may I know if all the overlapped variants are used in the training or is it down sampled?

Post edited by Geraldine_VdAuwera on

Best Answer

  • rpoplinrpoplin Posts: 122GATK Developer mod
    Answer ✓

    Hi atks,

    There isn't any downsampling of the training variants however there is some filtering that is done. By looking at the 1D annotation distributions, extreme outliers are removed from the training set. This behavior is controlled with the --stdThreshold argument to VariantRecalibrator.

    I hope that helps,

Sign In or Register to comment.