How am I supposed to write a process that takes multiple files, each of which has meta? Do I need to use a spreadsheet? I'm thinking of using the output of rmats/prep, which looks like

tuple val(meta), path("*.rmats"), emit: rmats

And I want to feed multiple rmats files to one rmats/post process. I guess I can use path and channels, but do I need a meta? The rmats/post needs to run on ALL input files in one process. Sorry - it feels like this should be obvious, but I've looked at the docs, and nothing seems to click

The output will be the path of all ".rmats" files processed from the given input. You either pass all expected inputs to generate the full list of rmats files, or you have to collect them afterwards. The requirement of "meta" depends on how the post process is defined. The output of

.collect()

would be a list of tuples [[meta1, file1.rmats], [meta2, file2.rmats], ...] or you can group by meta:

ch_rmats = RMATS_PREP.out.rmats
    .map { meta, rmats -> [ meta.contrast, rmats ] }
    .groupTuple()

where

meta.contrast

defines the grouping you want to apply for the

RMATS_POST

process

The #C0493FQQZ3P pipeline uses rMATS, maybe you can find something in their code that is helpful.