DepthOfCoverage rounding up to 100 percent
Dear GATK (aka. Dear Geraldine:))
I've recently experienced that DoC rounds off the "%_bases_above_30" up to 100% in the sample_summary output for one of our gene panels for certain samples, albeit a single exon is clearly not covered 100% at that -ct threshold. When rerunning DoC using only the relevant exon, the sample_summary correctly states the "%_bases_above_30" value. It's not a large genepanel (23 genes, ~70Kb ROI), so it seems like a bug that the overall (sample_summary) "%_bases_above_30" is calculated as 100% when the "%_bases_above_30" for one of the exons used for this calculation is only ~90%.
While that particular exon only constitutes ~0.2% of the entire ROI, it should never be possible to round up to 100%... My best guess is, that DoC sample_summary rounds of to 1 decimal (e.g 99.97 is rounded of to 100.0) without the constraint that rounding up to 100% should not be possible.
Could you clarify this?
Thanks in advance.