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.

Variant Annotator and other CALLERS

BogdanBogdan Palo Alto, CAMember ✭✭

Dear all,

talking about VariantAnnotator, could it be safely used on VCF files produced by other callers like STRELKA, VIRMID, Varscan2 etc ...beside MUTECT1 or MUTECT2 ? I am referring to the use of VariantAnnotator with the options for FS, SOR, RMSMappingQuality, QualByDepth, ReadPosRankSumTest ...

or, are there any minimal requirements for a VCF file produced by other callers in order to be fully annotated by VariantAnnotator ?

thank you,

-- bogdan

Comments

  • SheilaSheila Broad InstituteMember, Broadie, Moderator admin

    @Bogdan
    Hi Bogdan,

    I think if the VCF files from those tools validate with ValidateVariants, you should be able to use VariantAnnotator on them. However, there are no guarantees as I have seen users report issues with VCF files produced from other callers.

    -Sheila

  • BogdanBogdan Palo Alto, CAMember ✭✭

    Thank you Sheila ! May I add one more question please : shall I use a VCF file (from MUTECT, STRELKA, VIRMID , VARSCAN2 or any other CALLER), with SNV and INDELS, all these fields -- FS, SOR, RMSMappingQuality, QualByDepth, ReadPosRankSumTest - could be safely asked to be computed (in the sense that the tool VariantAnnotator has been tested successfully with these fields) ?

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie admin

    @Bogdan, these are not things that we test systematically because this is not how we use these tools. In general, if the annotation tool runs successfully, you can expect the annotated values to be correct relative to the bam files that were used. However you should be aware that comparing these annotations across callsets produced by different callers may not be an appropriate way to evaluate results of different callers, depending on what algorithms and internal data transformations they apply.

    We cannot advise you any further on these matters as they are outside the scope of support we provide; good luck.

  • BogdanBogdan Palo Alto, CAMember ✭✭

    Thanks Geraldine ....possibly if you do not mind, could we ask the MUTECT2 and VariantAnnotator development team about the other question I have posted : http://gatkforums.broadinstitute.org/gatk/discussion/7495/variantannotator-and-mutect2

    your answer would be extremely helpful in terms of deciding on how we move forward with my project ;).

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie admin

    I answered just now. Please note that we may have to delay answering some of your questions as you have been asking a large volume of questions recently. There are many people who need our help and we need to make sure we give everyone a fair chance of getting their questions answered. Specifically, for the next two weeks we may not answer any questions about doing things in a way that diverges significantly from the best practices or default tool behaviors.

  • BogdanBogdan Palo Alto, CAMember ✭✭

    thanks Geraldine, I am very grateful for your time and help ! many thanks also to Sheila !

Sign In or Register to comment.