Identification of MXE events

[Salwa Mostafa]

Hello,

I was wondering what the criteria is for identifying two exons as "mutually exclusive". For example, APH1A was identified as undergoing MXE in our dataset. The first exon is chr1:150267715-150267789 and second exon is chr1:150267956-150268127 (hg38 genome).

However, there are APH1A transcripts that include both or neither of these exons. So why were these exons identified as "mutually exclusive"?

Thank you!

Best,
Salwa.

[kutsc...@gmail.com]

rMATS requires that both the inclusion isoform and skipping isoform have evidence of the 3 exons in that isoform being sequential in a transcript. That transcript evidence could either be from an annotated transcript from the --gtf or it could be from some alignment from the input bam files. Here's the code: https://github.com/Xinglab/rmats-turbo/blob/v4.2.0/rMATS_pipeline/rmatspipeline/rmatspipeline.pyx#L1537

The diagram in the README shows the structure of the inclusion and skipping isoforms for MXE events: https://github.com/Xinglab/rmats-turbo/tree/v4.2.0#output

Basically if rMATS thinks that both the inclusion transcript and skipping transcript are real transcripts then rMATS detects an MXE event. The code doesn't require that the two exons are never found in the same transcript or that all transcripts have at least one of the exons. The criteria for rMATS to output an MXE event is not very strict which should result in most real MXE events being reported, but also many cases that are not really mutually exclusive

Eric

[Xiao Lei]

Hi, Eric,

I have a related question, in MXE cases, it is not unusual to see reads cross the junction of mutually exclusive exons (in your diagram, it will be e1 and e2 exons, image attached), how does rMATS deal with these reads?  Does rMATS just ignore these reads?

Thanks,

Xiao

[kutsc...@gmail.com]

The junction going from e1 to e2 is not part of the event. If a read includes that junction then rMATS will not count that read toward the event and rMATS essentially ignores that read for the MXE event. Here's the rMATS code for counting junction reads for MXE events: https://github.com/Xinglab/rmats-turbo/blob/v4.2.0/rMATS_pipeline/rmatspipeline/rmatspipeline.pyx#L2023

Eric

[X L]

Hi Eric,

Regarding the criteria for reporting MXE events: if rMATS requires that both the inclusion and skipping isoforms show evidence of all three exons being sequentially spliced in a transcript (as you mentioned), does that mean a read must be long enough to span the entire exon (either e1 or e2) along with its flanking junctions—that is, from upstreamEE to the exon and then to downstreamES—to support the event?

Best regards,
Xiao

On Thursday, December 28, 2023 at 4:07:27 PM UTC-5 kutsc...@gmail.com wrote:

[kutsc...@gmail.com]

The read support for a 3 exon sequence actually is not from a read with 2 junctions, but from a single junction of a read combined with an annotated transcript. See this issue: https://github.com/Xinglab/rmats-turbo/issues/512

Eric