Keep orphan read mappings in transcriptome BAM

10 views
Skip to first unread message

Daniel Fulop

unread,
Dec 22, 2021, 9:44:27 AM12/22/21
to rna-star
Hi Alex,

Thanks for the awesome and full-featured mapper!

I have a question regarding transcriptome BAMs. I am doing bulk RNAseq on bacteria and feeding STAR's transcriptome BAM into Salmon.

Because of bacteria's polycistronic mRNA I'd like to keep orphan read mappings in the transcriptome BAM, i.e. each read of a mate pair could somewhat often map to different but adjacent transcripts/genes and thus be considered orphan reads or singly mapped.

As far as I could tell from the documentation keeping orphan reads is not an option for the transcriptome BAM. Is this correct? Would outputting orphan reads in the transcriptome BAM constitute a new feature request?

Cheers,
Dan Fulop.

Alexander Dobin

unread,
Dec 27, 2021, 2:07:17 PM12/27/21
to rna-star
Hi Dan,

you are right, the alignments whose mates are concordant with different genes cannot be output to Transcriptome BAM.
A possible workaround is to re-map such reads as single-end, and then merge the results with the original paired-end mapping.

Cheers
Alex
Reply all
Reply to author
Forward
0 new messages