rmats-long on a single sample

[eltonjrv]

Hi there,

I wonder if there's a way to run rmats-long on a single sample, like it is doable through the --statoff option in rMATs-turbo for short reads.
I only have one ONT RNA-seq (cDNA) library and would like to investigate any new splice variants arising from some target loci that might be different from the reference annotation.

Thanks,
Elton

[EricKutschera]

When rMATS-turbo is run with a single sample and --statoff it detects splicing events and gives the read counts for both isoforms in each event

With long-reads, you could run ESPRESSO with a single sample and it would detect the isoforms and give the read count for each isoform

rMATS-long uses the ESPRESSO isoform counts and then detects differential isoform usage between two sample groups. Then for the identified differential isoforms it makes plots and determines the splicing differences. With only one sample rMATS-long won't be able to check for differential isoforms, but you could still use the scripts for plotting and checking splicing differences if you have some specific isoforms you want to check:
https://github.com/Xinglab/rMATS-long?tab=readme-ov-file#visualize-isoforms-and-abundance
https://github.com/Xinglab/rMATS-long?tab=readme-ov-file#classify-isoform-differences

Since you mentioned splice variants that differ from the reference annotation, you could try looking for novel isoforms identified by ESPRESSO that have a lot of assigned reads. Those novel isoforms will have transcript IDs that start with ESPRESSO:

[eltonjrv]

Thanks a lot, Eric.
I'm just looking into running ESPRESSO itself at the moment.
Best,
Elton