[miso-users] Problem with discarded reads from strand-specific library

[Andreas S. Richter]

Dear MISO users, 

I recently started using MISO with input data from strand-specific libraries. I configured MISO accordingly by setting "strand = fr-secondstrand". 

When inspecting the batch-logs generated by run_events_analysis.py, I noticed the following curious behaviour: 
* For the vast majority of events on the - strand, nearly all reads are discarded due to strand violation (example below) 
* For events on the + strand, only few reads are discarded by this filter. 

I visually inspected the reads of the BAM file together with the annotation in IGV and both looked correct regarding strandedness. 

Can someone please explain me this behaviour? I am using version 2 of mm10 alternative events and MISO version 0.4.9. 

Thanks a lot in advance. 

Best, 
Andreas 


Example output of run_events_analysis.py: 
Computing Psi for 1 genes... 
- 1:97661831:97662018:- at 1:97654600:97654689:- at 1:97643902:97646361:- 
... 
Filtered out 61 read pairs that were on same strand. 
Filtered out 60 reads that had no paired mate. 
- Total read pairs: 2458 
No. reads discarded due to strand violation: 2458 
Only 0 reads in gene, skipping (needed >= 20 reads) 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mit.edu/pipermail/miso-users/attachments/20130807/92c30114/attachment.htm