fromGTF.*.txt files have different events from sample to sample

[mblanche]

Hi,

I'm curious as to why to different samples using the same GTF file have different content in the fromGTF.*.txt files. I was on the impression that this was the ensemble of alternative splicing event existing in the GTF universe.

Thanks

[EricKutschera]

The fromGTF.[AS_Event].txt files have all the detected events which could include events with a novel junction seen in the alignments (fromGTF.novelJunction.[AS_Event].txt) or a novel splice site (fromGTF.novelSpliceSite.[AS_Event].txt): https://github.com/Xinglab/rmats-turbo/tree/v4.3.0?tab=readme-ov-file#output

If you want to find the set of events that rMATS can detect from just the GTF you can run with the empty.bam from this post: https://github.com/Xinglab/rmats-turbo/issues/79

[mblanche]

Thanks Eric for the fast response.

So, just to make sure I understand completely, if an event exists in the GTF but is not detected in the BAM files of a comparison set, this event will NOT make it into the fromGTF.[AS_Event].txt. For some reason, reading the docs, I thought that the fromGTF.[AS_Event].txt was the ensemble of all possible event in a genome, present of not in the bam files...

[EricKutschera]

The fromGTF.[AS_Event].txt file will include all the events (that rMATS can detect) that exist in the GTF. The fromGTF file will also include events that were detected based on novel junctions or splice sites that have supporting reads in the BAM files. So the fromGTF file contains events from the GTF plus some extra events based on the BAM files