[miso-users] About novel events not in the MISO GFF3 annotation.

[Tang, Xiaojia, Ph.D.]

Hi everyone on the miso ML,

I have been using MISO for identifying possible alternative splicing events in RNA-Seq data. It works well for events that are in the GFF3 files. However, this time in a new analysis we are expected to detect some exon-skipped events in a known isoform, which would result in a novel isoform. We have been able to see it happen in IGV. But we will need to use some algorithm in order to detect such events genome-widely. I know that MISO's event-centric or isoform-centric model would work basing on the provided events, and also only detected those events/isoforms. Is there a way to detect possible skipping events of a known exon which might be a novel event resulting in a novel isoform?

Thanks,
Xiaojia
Informatics Specialist
Department of Health Sciences Research
Mayo Clinic Rochester
Rochester, MN 55901

[Yarden Katz]

Hi Xiaojia,

MISO doesn't do novel isoform detection, only quantitates isoforms based on the annotation it is fed, as you noted. One thing some of our users have done is to run a de-novo isoform assembly program (or use some other means to compile an annotation with novel isoforms), convert these de novo isoforms into events in GFF3 format, and run MISO on the resulting annotation to get isoform expression.

Yarden

Post by Tang, Xiaojia, Ph.D.
Hi everyone on the miso ML,
I have been using MISO for identifying possible alternative splicing events in RNA-Seq data. It works well for events that are in the GFF3 files. However, this time in a new analysis we are expected to detect some exon-skipped events in a known isoform, which would result in a novel isoform. We have been able to see it happen in IGV. But we will need to use some algorithm in order to detect such events genome-widely. I know that MISO’s event-centric or isoform-centric model would work basing on the provided events, and also only detected those events/isoforms. Is there a way to detect possible skipping events of a known exon which might be a novel event resulting in a novel isoform?
Thanks,
Xiaojia
Informatics Specialist
Department of Health Sciences Research
Mayo Clinic Rochester
Rochester, MN 55901
_______________________________________________
miso-users mailing list
http://mailman.mit.edu/mailman/listinfo/miso-users