[miso-users] Why are exon-centric events always involving two isoforms?

[Yarden Katz]

Further clarification: exon-centric as we use it generally means two isoforms, but it doesn't have to.  If you have an exon that is alternatively skipped, but also contains an alternative splice site to one of its flanking exons, you can choose to look at this as a multi-isoform event containing 4 isoforms.  You can build that into your own GFF annotation.  Then when you run MISO for that event, you'd get a vector of 4 Psi values, one for each isoform, instead of just a single number. 

  --Yarden

On Jan 23, 2013, at 10:39 PM, Yarden Katz wrote:

> Many splicing events can be more complex, as you say.  The decision to focus on two isoforms is mainly for convenience, ease of (biological) validation, and interpretability.  You don't have to use a two-isoforms only annotation -- MISO can take any GFF annotation as input, including ones that describe every transcript per gene, so in that case you'd get a Psi value for each of the multiple isoforms of a gene.  Quantification of multiple isoforms is harder to do since RNA-Seq are inherently relatively short, and annotations are in general imperfect/incomplete, and are also harder to analyze, but might be more appropriate depending on your analysis.
> 
> In the two isoforms approach, events are split up into pairs from more complex event -- we used the annotation from Wang et. al. (2008) (see http://genes.mit.edu/burgelab/miso/docs/#alternative-event-annotations) except for TandemUTR events, where we used polyA DB. 
> 
>  --Yarden
> 
> On Jan 23, 2013, at 10:27 PM, Li Shen wrote:
> 
>> Although I know this is rare, could events like AFE have three or even
>> more alternative TSS? How does MISO handle this? Split them into two
>> events?
>> 
>> --
>> Li Shen, Ph.D. Computer Science
>> Assistant Professor
>> Department of Neuroscience
>> Mount Sinai School of Medicine
>> One Gustave L. Levy Place, Box 1065
>> New York, NY 10029
>> Phone: 212-659-8622 (office)
>> _______________________________________________
>> miso-users mailing list
>> miso-users at mit.edu
>> http://mailman.mit.edu/mailman/listinfo/miso-users
> 
> 
> _______________________________________________
> miso-users mailing list
> miso-users at mit.edu
> http://mailman.mit.edu/mailman/listinfo/miso-users