haplotypeCaller phasing
Answered
Hi,
Does HaplotypeCaller use paired end information (considering paired-ends as a single read), for its phasing?
Best,
-
No, HaplotypeCaller does not use paired end information for phasing. Mutect2 does use this information though.
-
mutect2 also outputs germline mutations in addition to somatic mutations. Why would one go on to use haplotype caller then? What are the advantages / differences?
-
If you are looking for germline calls, I would recommend HaplotypeCaller. HaplotypeCaller is meant to call germline mutations, the germline output from Mutect2 are not always as reliable.
-
Thank you for your answer!
"the germline output from Mutect2 are not always as reliable."
Is there a good reason as to why, though?
-
There have been some good discussions in past forum posts covering why we recommend getting germline calls from HaplotypeCaller. Here are the links:
- Call germline mutations from Mutect2
- Germline sites in Mutect2
- Germline mutations in Mutect2 vs HaplotypeCaller
We also have a couple articles that might be helpful in your decision, which is the 'Somatic calling is NOT simply a difference between two callsets' and 'FAQ for Mutect2'.
Please let me know if you have further questions.
-
Oh thanks! these are some great ressources.
From what I see, I will use M2 with the --genotype-germline-sites set to true in order to maximize variants found and filter found germlines based on AF.it might still be somewhat useful to use HC and compare the germline variants that are found by both and if there is important differences.
Going back to two things then:
1. Per David's comment:
> G E Turning on -genotype-gemline-sites should not affect downstream filtering, but unfortunately I neglected to exclude germline sites from the clustered events filter, and so turning on this argument, thereby emitted more germline variants, makes the clustered events filter slightly overactive. Therefore I can only recommend this argument for diagnostic purposes even thought the effect is small.
It might be nice to exclude germline sites from this filter by default? How would I do that myself?
2. there is no way to set the --genotype-germline-sites filter from the GATK's M2 WDL workflow. It would be nice to add this as a parameter for the workflow
Thanks again for the help! -
Great! I'm glad they were helpful.
1) There wouldn't be a workaround to exclude germline sites from the clustered events filter.
2) Could you point to which specific WDL you are referring to? I'll put in a feature request.
-
You can currently use that option by adding to the m2_extra_args input. Here is an example in the sample inputs json: https://github.com/broadinstitute/gatk/blob/82302587fff9aed51060bfe3258cbab16d4d82de/scripts/m2_cromwell_tests/mutect2.inputs.json#L6
-
of course.. my bad ahah. Thanks again!
-
No problem, glad we could find a solution!
Have a good one.
Please sign in to leave a comment.
11 comments