I encountered an issue with the bowtie rrna remova...
# riboseq
j
I encountered an issue with the bowtie rrna removal due to duplicate sequence naming (or at least what some tools interpreted as such). Anyone here care to review the fix? https://github.com/nf-core/modules/pull/9522
✅ 1
j
Looks good to me!
Did you ever think subworkflows would get this complex.. Is there any sense in this being multiple different subworkflows rather than all in the one and people do their selection on adding the subworkflow rather than in config
j
I think this is useful independently of the higher-level subworkflow, so it makes sense for it to be factored out
Thanks for the review!
j
Oh no doubt. I meant more that there would be multiple different subworkflows with slightly different strategies available through nf-core/modules. Or do you see it similar to the initial philosophy around pipelines whereby fragmentation is bad
j
Could you explain more what you mean with an example?
j
Its likely not worth your time. What I was wondering if it makes more sense to have different subworkflows that support each of the removal tools rather than one subworkflow that will increase in complexity each time a new tool, with its own requirements, is added
j
Ahh, well this subworkflow started out with its purpose being pretty much entirely to allow a single import and selection of removal method, there wasn't much to any of the pathways. The bowtie2 pathway has just extended a little for the preprocessing! There's no hard and fast rule for this, this is mainly a consequence of the way I think, but this way would allow us to share any common preprocessing required, for example, rather than replicating it for each of the tools (though it's not currently the case).