Using LearnReadOrientationModel produces new mutations?

Follow

vctrymao

• May 11, 2020 22:55

Using GATK 4.1.2.0. I included the new LearnReadOrientationModel into my Mutect2 calling pipeline, and I noticed that while some additional calls were filtered out, some new mutations were added in. Is this possible, and why is this possible?

0

• 

  Bhanu Gandham

  • May 11, 2020 23:50

  Hi ,

  The GATK support team is focused on resolving questions about GATK tool-specific errors and abnormal results from the tools. For all other questions, such as this one, we are building a backlog to work through when we have the capacity.

  Please continue to post your questions because we will be mining them for improvements to documentation, resources, and tools.

  We cannot guarantee a reply, however, we ask other community members to help out if you know the answer.

  For context, check out our support policy.

   

  0

  Comment actions Permalink
• 

  David Benjamin

  • May 12, 2020 01:23

  vctrymao This is possible.  FilterMutectCalls sets a threshold on the posterior probability that variants are real by optimizing the estimated F score, which is the harmonic mean of sensitivity and precision.  When variants are correctly filtered out as orientation bias artifacts, FilterMutectCalls may nudge the threshold a bit lower in order to get a bit more sensitivity.  I would hope that all of the new PASS variants are borderline calls.

  0

  Comment actions Permalink

Please sign in to leave a comment.