bam-readcount values are higher than Mutect results

Question

[sey]

I use Mutect to find somatic mutations in multiple tumor samples (to be used in clonality analysis). I don't get the readcounts if the caller did not identify a somatic mutation in a sample. I'm using bam-readcount to fill in the empty values after merging Mutect results from different samples. Where I have values from both tools, the readcounts are lower in the Mutect results. Is there a way to use bam-readcount in a way to substitute for missing Mutect values?

Answer 1

[Chris Miller]

Many callers do some filtering of reads first (by base quality, by mapping quality, maybe even local realignment) and report only counts from reads that pass. bam-readcount, by default, outputs "raw" counts. You may or may not be able to mimic those Mutect settings with bam-readcount parameters.

This problem is compounded when merging the output from different variant callers, and there is no "perfect" solution.