Hello,
I typically use RSEM and strange issues like this do not arise in the first place. However, for one particular (very narrow-scope) task, transcriptome alignment is not acceptable and old good HTSeq comes into play. The problem is in the reads that map to multiple locations. We do need those (counted once, and we don't care about the particular feature they map to; we are talking about duplicated rRNA features and a small bactrial genome). The reads of interest are, of course, discarded by HTSeq, and I am puzzled how to force HTSeq to count them, by hacking either HTSeq or bam files. The dropped reads have non-empty XA tag AND zero mapping quality (that's what I see in IGV).
Any smart suggestions?
Thanks!
I typically use RSEM and strange issues like this do not arise in the first place. However, for one particular (very narrow-scope) task, transcriptome alignment is not acceptable and old good HTSeq comes into play. The problem is in the reads that map to multiple locations. We do need those (counted once, and we don't care about the particular feature they map to; we are talking about duplicated rRNA features and a small bactrial genome). The reads of interest are, of course, discarded by HTSeq, and I am puzzled how to force HTSeq to count them, by hacking either HTSeq or bam files. The dropped reads have non-empty XA tag AND zero mapping quality (that's what I see in IGV).
Any smart suggestions?
Thanks!
Comment