[miso-users] Unannotated SE events and multi-mapping reads

[Yarden Katz]

Hi Sahin,

See comments below:


On Jul 21, 2013, at 5:02 PM, Sahin Naqvi wrote:

> Hi,
> 
> I have two questions about MISO:
> 
> 1) I'm interested in quantifying the PSIs of the exons of a few genes of interest, and have noticed that the annotated skipped exon events only contain a fraction of the possible skipped exon events (which makes sense given that they are from existing annotations). How would MISO deal with reads that were consistent with some other SE event? If it ignores that, is there a way for me to manually add potential events to the alternative event annotations?

Reads that are inconsistent with the annotation will be discarded.  It's easy to add your own annotations, however: just add them to the GFF file and re-index it (with 'index_gff.py') and MISO will run it like any other event.  See this for details on the format of the GFF:

http://genes.mit.edu/burgelab/miso/docs/#gff-based-alternative-events-format

Note that the IDs of the GFF are not parsed by MISO except to find the hierarchy of gene -> mRNA -> exon, so you can format your IDs however you like (the elaborate MISO ID values for the annotations we compiled are arbitrary). 

> 
> 2) How does MISo deal with multi-mapping reads when quantifying PSI values? Is it similar to the cufflinks method?

It currently doesn't do anything special for multi-mapping reads.  If a read maps to two distinct genes/events and you allow that in your BAM, it'll use those reads in each gene/event like uniquely mapping reads.

Best, --Yarden


> 
> Thanks,
> Sahin Naqvi_______________________________________________
> miso-users mailing list
> miso-users at mit.edu
> http://mailman.mit.edu/mailman/listinfo/miso-users