Hi. I've been tackling the PR to get rmats into nf-core. https://github.com/nf-core/modules/issues/8699 My planned approach is to have multiple tasks • RMATS prep • RMATS post • statistical setup, which would be a bit like contrast reading in DESEq2 • RMATS stats That way, RMATS prep can run on each BAM file separately, and multiple comparisons can be done (regular RMATS is used to just comparing two groups, but if you want A vs B, A vs C you need to separate all the tasks). I will have one subworkflow that runs all the steps, so it can be used in a more convenient manner. My two questions are 1. How should go about making PRs for this. Right now, I've mostly finished RMATS prep. Should I make one PR for each stage (4 tasks, one subworkflow), or should I make one big PR? 2. Are there any problems with this overall approach?

Making individual PRs is easier to review (and fix) and generally advised, especially with 4 modules + subworkflow. Are the modules clear tools from RMATS (I have no idea what this tool does)

Rmats is one program for splicing that has 3 commands test can be given to the same script rmats.py Prep Post Stats They can be run together on all samples, but can be run alone. Prep needs BAM files the rest just need txt files outputted by the previous step. There are helper scripts to set up stats, which i plan to have in their own task (need to review the scrpts). Does this make sense?

Yep, that makes sense. Just thinking about what the module names should be, but as they directly have the commands in the tool that makes it really simple. I think that is clear, do you need anything else clearing up?

I'm good for now. I'll have more questions when I get to the actual specific modules and the workflow. Thank you!