falco doesn't consider BAM flag bits

[wm75]

It seems that falco doesn't handle reverse-complemented read sequences in mapped BAM files correctly.
Specifically, when flag bit 0x04 of a read is not set (i.e. when the read is mapped), falco should consider the status of the 0x10 bit and reverse-complement the sequence if that bit is set.

Example output of falco (multiqc visualized) on just the reverse-strand reads from a mapped bam:

and those same reads analyzed with fastqc:

[bgruening]

@andrewdavidsmith this is blocking us from deploying falco as a real alternative for fastqc. Can you please have a look?

Thanks!

[andrewdavidsmith]

Yes, thanks. I can take a look within the next couple days.

[andrewdavidsmith]

@bgruening I have a fix in branch bam-strand-fix that shows these problems corrected for per-position quality scores and per-position nucleotide distributions. If you can check by pulling and building that specific branch it would be great. I should be able to merge this into the master branch soon.

[wm75]

I tested the version on that branch and it behaves just like fastqc now.
Looking forward to seeing this in a release and thank you!