Notice:
If you happen to see a question you know the answer to, please do chime in and help your fellow community members. We encourage our fourm members to be more involved, jump in and help out your fellow researchers with their questions. GATK forum is a community forum and helping each other with using GATK tools and research is the cornerstone of our success as a genomics research community.We appreciate your help!

Test-drive the GATK tools and Best Practices pipelines on Terra


Check out this blog post to learn how you can get started with GATK and try out the pipelines in preconfigured workspaces (with a user-friendly interface!) without having to install anything.

GenotypeGVCFs 3.7 gives different results depending on sharding

Broad's BP joint calling workflow shards GenotypeGVCFs by genomic locations. I was tasked to port the BP workflow from GATK4 to GATK3.7, and reference genome 37.

https://github.com/gatk-workflows/gatk4-germline-snps-indels/blob/b6f86ebfeeb8990b2588b72f3ad869e93e218e5f/joint-discovery-gatk4-local.wdl#L115

I was not able to find an equivalent of the hg38.even.handcurated.20k.intervals file for hg37, so tried to make one myself. (I believe people have asked before on this forum for hg38 versions of these intervals, or an explanation of how they were created).

I found that depending where the intervals lie, different calls are emitted by GenotypeGVCFs. The first difference I have seen is that ALT '*' alleles are not called if the upstream deletion is outside the interval.

Called with these shards:

21:10705896-10707251
21:10707252-10707407
21:10707408-10722075

21 10707250 . CTTTGTGA C 140.17 . AC=1;AF=0.0005097;AN=1962;BaseQRankSum=2.25;ClippingRankSum=0;DP=1
21 10707253 rs71227073 T A,* 15301.9 . AC=96,1;AF=0.049,0.0005097;AN=1962;BaseQRankSum=1.62;Clipping...


VS


21 10707250 . CTTTGTGA C 140.17 . AC=1;AF=0.0005097;AN=1962;BaseQRankSum=2.25;ClippingRankSum=0;DP=1
21 10707253 rs71227073 T A 15301.9 . AC=96;AF=0.049;AN=1962;BaseQRankSum=1.62;ClippingRankSum=0;DB;D

Is this expected behaviour? IE that calls near the shard boundaries are incorrect?

Is there documentation about the method used to create hg38.even.handcurated.20k.intervals ?

Is there a way to workaround this issue?

Thanks

Best Answer

Answers

Sign In or Register to comment.