Extracting descriptive metrics from variant reads in BAM

Question

[ognjen011]

Hello all!

I would like to extract the following descriptive metrics from BAM:

• mean base quality
• mapping quality
• mean position in read for variant reads

I would like to go through a VCF and find those metrics for variant-supporting reads directly from BAM. I can do it using pysam, but I was wondering if there is an out-of-the-box solution that is ready and perhaps computationally more efficient. Any help is welcome!

Thanks in advance!

P.S. I also know that VarDict Java does it for candidate locations, but sometimes I need it for locations not recognized by VarDict's algorithm. But in general, such an output (TSV) would be perfect.

Answer 1

[Pierre Lindenbaum]

I think GATK variant annotator is able to do this: https://software.broadinstitute.org/gatk/gatkdocs/3.8-0/org_broadinstitute_gatk_tools_walkers_annotator_VariantAnnotator.php

There are a bunch of predefined annotations under: "Annotation Modules" https://software.broadinstitute.org/gatk/gatkdocs/3.8-0/org_broadinstitute_gatk_tools_walkers_annotator_VariantAnnotator.php

mean base quality

https://software.broadinstitute.org/gatk/gatkdocs/3.8-0/org_broadinstitute_gatk_tools_walkers_annotator_BaseQualityRankSumTest.php

mapping quality

https://software.broadinstitute.org/gatk/gatkdocs/3.8-0/org_broadinstitute_gatk_tools_walkers_annotator_AS_MappingQualityRankSumTest.php

mean position in read for variant reads

https://software.broadinstitute.org/gatk/gatkdocs/3.8-0/org_broadinstitute_gatk_tools_walkers_annotator_ReadPosRankSumTest.php

[ognjen011]

Thanks for the advice! Many of these are actually very useful, but I still wonder - is there a way to get mean BQ, MQ and position, and not the results of the rank sum test?