Entering edit mode
6 weeks ago
STARDUST • 0
I have ploy-A selected RNA library for wild-type with kock-down of the protein of interest. I managed to do DESeq and DEXSeq. DESeq results are very straight forward but DEXSeq bit confusing for me. After DEXSeq, i have around ~4000 exons. My next task is to find what is the most frequent alternative-splicing event from DEXSEQ data. How can find that. Can i use DEXSEQ data or any alternatives?
What do you mean be "most frequent alternative splicing event"? Do you mean which type of change (cassette exon, alternative 5' splice-site, intron retention etc) is most common amongst the significant changes, or do you mean "which exon is most often changed between normal and treatment?"
To be honest, i wanted to know only which type of change. since you mentioned about the specific exons, if it is possible i wanted to know which exon affected.
DEXSeq is probably not the correct tool for what you are trying to do. If focuses of changes to which exons are included and which excluded. It's obvious how this can be applied to cassette exons. It can be applied to other sorts of changes, but it's more complicated, both conceptually and technically. Instead i could try something like rMATS or SUPPA if you are interested in categorising changes.
I will try both rMATS and SUPPA for categorising changes. For the specific exons, i'm expecting the last exons should be affected. That is our one of the hypothesis. What i have in my mind is that i will take all the transcripts that exons are differentially used. And calculating RPM/TMM values and doing simple metaplot like that. But i don't know how to take care of the exons that orvelap's with other transcripts.