We've moved!
This site is now read-only. You can find our new documentation site and support forum for posting questions here.
Be sure to read our welcome blog!

Recommended Pipeline

mike_boursnellmike_boursnell Member
edited July 2012 in Ask the GATK team

Hi. Could you suggest a recommended pipeline for us. We receive BAM files from our sequencing - the sequencing is mostly targetted regions (~10MB) and the species is dog, so we don't want to skip known SNPs as different breeds have different known SNPs. We have one BAM file for each dog sample, usually about 10 samples.

I have looked at the guidelines (Best Practice Variant Detection) for this but I'm a bit confused...


  • SophiaSophia Member

    After that, to obtain all samples in one vcf file, you might run the UnifiedGenotyper analysis listing all processed bam files in the same command.

  • SophiaSophia Member
    edited July 2012

    PS.: GATK does not know about dog SNPs and will not consider or skip them unless you tell it to (e.g. by inputing a vcf or similar positions file in some of the commands). That means you will get unfiltered results to start with.

  • Thanks for those comments.

    In the Best Practice Variant Detection it has these options:

    • Fast: lane-level realignment at known sites only and lane-level recalibration
    • Fast + sample-level realignment
    • Better: sample-level realignment with known indels and recalibration
    • Best: multi-sample realignment with known sites and recalibration

    Are you suggesting using the last of these options? (Best...)

    Can this be done through the DataProcessingPipeline Queue script?

  • SophiaSophia Member

    Mike, do you have lists of known SNPs from dog? Which dog reference genome version are you using?

  • Hi Sophia. We have a list of known SNPs from canFam3. Do you have a list as well? If so I wouldn't mind comparing it to ours. We don't have a list of known Indels for the IndelRealigner though.

Sign In or Register to comment.