STAR diploid transform for insertions about star

[alexdobin]

Hi Matt,

If you do not see the reads, it means they were classified as unmapped. They will not be counted in quantification.

[alexdobin]

Hi @mparker2

Such reads should be reported as unmapped, but I have never checked that - if you can select and check them, it would be great.

[mparker2]

Hi @alexdobin,

Thanks for the response. I found some alignments that just had 1 matching reference base (reported as 149S1M or 1M149S), but no 0-length alignments:
I identified some reads that should map to this 2365 bp indel. None of them were reported by STAR diploid (I didn't use --outSAMunmapped Within). I would have to realign the data and keep unmapped reads to see if they still have the ha tag. Would they also be ignored in quantification (the indel falls within an annotated gene)?

[mparker2]

Circling back to this issue (sorry)... I've noted that some tags such as CB are retained for unmapped reads, which is very useful. It would be similarly useful to retain ha:i:2 tags for reads that map to the second haplotype but cannot be transformed into haplotype 1 coordinate space. Would this be possible to do?

[alexdobin]

To see the alignments of such reads, you can run mapping without the back transformation.