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.
We will be out of the office on November 11th and 13th 2019, due to the U.S. holiday(Veteran's day) and due to a team event(Nov 13th). We will return to monitoring the GATK forum on November 12th and 14th respectively. Thank you for your patience.

MuTect2 beta --germline_resource for build h19,af-only-gnomad.vcf

Hi - I'm looking to run MuTect2 beta using the --germline_resource option. However,I cannot find af-only-gnomad.vcf for build h19.
How can I find this vcf for hg19?

Best Answer


  • jemimalwhjemimalwh ChinaMember

    @Geraldine_VdAuwera said:
    We do not provide it. You can try converting it using Picard LiftoverVcf or the UCSC liftover web utility.

    Hi Geraldine. I came to your talks in BGI few days ago and found them very useful. I have a few questions when I test Mutect2:

    1. I am doing clinical analysis, and in most of the case my data belongs to one individual. So I don't have a batch of Normal samples to create PoN. I wonder what effect/bias will happen if missing this step.

    2. Also I am trying to use the parameter "--germline-resource", but I found the af-only-gnomad.raw.sites.b37.vcf.gz file (while I use hg19; downloaded from includes sites only belong to chromosome 1. I wounder if this is a complete version, or I should downloaded this dataset by myself? (

    Thank you.

  • SheilaSheila Broad InstituteMember, Broadie admin
    edited April 2018


    1) If you don't have a PoN, you may get more false positives that are sequencing artifacts. If possible, you can try to find publicly available data to use in your PoN. Have a look at this article for more information. Also note, having a matched normal and germline resource will help a lot with filtering out potential germline variants.

    2) You can try to do a liftover, like Geraldine mentioned in the above post.


  • escaonescaon Limoges, FranceMember

    Regarding the "af-only-gnomad.raw.sites.hg19.vcf.gz" file :

    1) I did browse the gnomAD google cloud but failed to find such a file.

    2) I might be doing a liftover, trying to convert af-only-gnomad.raw.sites.b37.vcf.gz to af-only-gnomad.raw.sites.hg19.vcf.gz using the b37tohg19 chain.

    3) Or i might "simply" use the af-only-gnomad.raw.sites.hg19.vcf.gz found HERE. But I'm unsure if I can trust this source.

    Was I supposed to find the file using 1) ?
    What do you recommand to use, 2) or 3) ?

  • SheilaSheila Broad InstituteMember, Broadie admin


    We do not provide the hg19 version, but we provide the b37 version you can liftover to hg19 here.

    You can try using option 3, but I have never tested it myself.


  • 29043594952904359495 Member
    edited March 28
    If you can liftover and put it there will be much better, we believe you than ourselves

    since you do not test the af-only-gnomad.raw.sites.hg19.vcf.gz, why here supply the file, a little confused,
    and why b37 version add a the word raw, but hg38 no this word, is there any hidden meaning

    since b37 and hg19 both are raw, why we can use b37 to conver, but not hg19
    thanks a lot
    Post edited by 2904359495 on
  • 29043594952904359495 Member

    The file is too big to load to ucsc browser to liftover, and I think the liftover file is not right, because variant in one version has hihg possibility not being a variant, but we expect a high confidence, so liftover is not right , I think

Sign In or Register to comment.