It looks like you're new here. If you want to get involved, click one of these buttons!
What are the BD and BI flags that get added to my bam files after base recalibration? They seem to consist of a long string of "N"s, and I'm trying to understand if that is correct.
Those are insertion/deletion base qualities, where N=45, which is the default quality score for indels assigned by the BaseRecalibrator. So it looks like your reads are not getting recalibrated and are just maintaining the default original qualities.
Geraldine Van der Auwera, PhD
Thanks. Is this expected behavior, or does it mean that something is wrong with the recalibration?
If you're only getting Ns everywhere, there may be something wrong, because nothing is getting recalibrated. What's you're command line?
I've looked at the quality strings and they do actually get changed, the few I spot checked mostly had individual characters decreased by an average of 5.
Looking at the command line, it seems that the --disable_indel_quals must be responsible. I'm not sure why I included this originally.
How are the BD and BI tags used downstream? Can I safely remove them if I am not using the UnifiedGenotyper to do genotyping?
Hi, I have a similar question to this. How are those tags now used with the newest version of GATK? Are they really helping for INDELs detection?
Thanks a lot
@JCGrenier The indel qualities are taken into account by the variant callers (specifically, the pairHMM algorithm) when assessing the amount of evidence there is for a particular variant, in much the same way that base qualities are used to assess evidence for SNPs.
Thanks Geraldine. Is there any quick way to add again those tags if I removed them from the racalibrated file? Or should I reprocess my files from a step before the recalibration?
Thanks for your help!
If you still have the recalibration tables you can just rerun the PrintReads step to generate the recalibrated bam.