Phred 33 and phred 64 scores mixed in .bam file

Question

[venks]

Hi,

I have a sample in my .BAM file which has phred score ranging from 1-66. The quality score distribution looks like there are two scoring system (phred 33 and 64 ) mixed when someone did alignment earlier.

I was wondering how this might have happened? Anyway of processing the BAM files to bring it down to phred 33 scoring system ?

Note: I have used --fix_misencoded_quality_scores -fixMisencodedQuals in GATK but this just subtracts 31 which makes my quality score negative.

Thank You very much

[Devon Ryan]

That's a first. Both SAM and BAM files should only ever have Phred+33, so I wonder how this was made. Do you know anything about that?

[venks]

Ryan,

Thanks for the reply.

The data is pretty old. It came from SOLID sequencing. It was later aligned using Novoalign.

[Devon Ryan]

Nothing says fun like trying to make sense of old data :(

[Matteo Schiavinato]

I am having the same problem in this moment, did you manage to solve it? If yes, how?