A community effort to collect a curated set of analysis pipelines built using Nextflow.

nf-core

what do you all think: is this calibration call inside a module okay, or is this against the atomic nature of modules? <https://github.com/nf-core/modules/pull/12359/changes#diff-80c27276027aa5752623c4100c944d066e58f465bfb8a504e504f75dbb553d65R100-R108>

I'm a bit confused because the docs you linked him actually imply he *should* have those meta maps for all except the value channel to me?

But in answer to your question here, I think the nf-core way would be to split them up. If he was streaming between the steps I think there would a stronger argument to justify grouping them? 

it is just weird that you would call the same module twice after each with more or less the same inputs

Does the calibration step not generate some inputs for the final step though? 

it does, that seems to be the only difference then

and yes, we generally have a meta map on everything at the moment