Dissecting number of forward/reverse reads from GATK output

Question

[newDNASeqer]

We use GATK to make variant callings on exome-seq dataset, and recently we need to know the number of forward and reverse reads that cover the mutations identified by GATK. However, GATK's output (vcf) does not contain such information.

I am thinking about using the mutation position as reported in the GATK output (e.g. chr, position) to walk through the BAM files so that I can count the reads myself. Before I do this, I would like to know if there's any existing tool for this purpose? Thanks

Answer 1

[vdauwera]

You can use the GATK's VariantAnnotator to add the Strand Bias (SB) annotation which will give you exactly that. Or request the annotation during the variant calling run. See the GATK documentation and support forum for more details.

[newDNASeqer]

when I try to use VariantAnnotator, I got the following error. I guess it no longer works with SB annotation. Looks like I'll have to re-run the HaplotypeCaller. I am using GATK 3.2

StrandBiasBySample can only be used with likelihood based annotations in the HaplotypeCaller

Answer 2

[Raony Guimarães]

Take a look

http://gatkforums.broadinstitute.org/discussion/4380/how-to-get-the-number-of-forward-reverse-reads-in-the-final-vcf-output

##INFO=<ID=SB,Number=1,Type=Float,Description="Strand Bias">

http://gatkforums.broadinstitute.org/discussion/1268/how-should-i-interpret-vcf-files-produced-by-the-gatk

SB How much evidence is there for Strand Bias (the variation being seen on only the forward or only the reverse strand) in the reads? Higher SB values denote more bias (and therefore are more likely to indicate false positive calls).