There is a lot of overlap, especially conceptually, but also in terms of the actual tools used in the two pipelines. In your 1, we don't yet have sortMeRna or similar, but it's on the issue list. In number 2, we have rnaSpades, but not Trinity, plus Megahit for assembly. In 3, we quantify with BBMap, but I've been thinking of adding Salmon. Of the quality tools in number 3, we only have transrate. Most of the tools in your 4, are also present in metatdenovo, although we have two prokaryotic ORF callers in addition to Transdecoder. We don't have Infernal nor Uniprot/Diamond and not Trinotate either. Many of the tools you have are hence things we wouldn't mind, so some common subworkflows would make a lot of sense to me. I don't know how to "cut" this best though, which granularity of swfs etc., so that's something we'd need to discuss. In the end, it could turn out as virtually the same pipeline with different names, serving to guide users to the right one for them, perhaps?

I will need to run both pipelines to get accustomed with their data and results. The first step is ofc to create all missing nf-core modules. I saw that Transdecoder is already nf-cored, but not used in any nf-core pipeline. Are you using a local implementation instead? Another task is to arrange a meeting with the original developer of #C04CWMEMWD7 to prioritize the updates; it would be awesome if you could join that one, so I will let you know if and when.

I'd be happy to join. (I wonder why we're not using the nf-core module for Transdecoder. If it didn't exist when we needed it we would have contributed one. Perhaps there's some quirk in the implementation we didn't like.)