Service notice: Several of our team members are on vacation so service will be slow through at least July 13th, possibly longer depending on how much backlog accumulates during that time. This means that for a while it may take us more time than usual to answer your questions. Thank you for your patience.

Seeing multiple calls for the same position


I am running a trio through the best practices pipeline for 3.7 and finding an unexpected result. I am seeing some positions that are being called multiple times with different calls. First, as output from HaplotypeCaller, I am seeing these results for position 78790153:

In each sample, the position 78790153 is represented by at least two lines, it's own line and at least one deletion prior. Of particular concern is that in the bottom case, the call at this position varies depending on which line you look at. It is heterozygous for a deletion at one point and homozygous reference at the others.

After Joint Genotyping, the variants come through in the result with that same position appearing twice still. (Order of samples top to bottom above is order of samples listed below first, second, third)

But, a new concern arises after applying family priors:

Here we see the first sample has her call from the deletion beginning at 78790143 changed from reference to heterozygous with no supporting evidence. This is obviously a very challenging spot because of the repeat, but is it deliberate for a position to be called multiple times and with different called alleles?




  • SheilaSheila Broad InstituteMember, Broadie, Moderator

    Hi Scott,

    Can you try with GATK4 or a later version of GATK3? I have not heard about HaplotypeCaller outputting sites twice, but in the Genotype Refinement Workflow, someone reported this issue, and the team is working on fixing it in GATK4. The fix will be in the next release.


Sign In or Register to comment.