OK guys, tricky one: what to do about having an 'nf-core workflow for proteomics'? quantms leaving creates difficulties for anyone wanting to contribute to similar in nf-core. We have collaborators who want to see something in nf-core, what would be the way to make that happen? Is it just a case of starting a new pipeline? How do we do it without starting a flame war, given that lots of the modules etc are likely to be similar? Seqera can't be on the hook for long-term maintenance of such a thing, do we have any folks who would be lead devs? @Jose Espinosa-Carrasco ?

I am afraid I am not a proteomics guy, we (ping @Júlia Mir) know some people at the CRG working on this field (proteomics unit) and they are developing #C08EKHU8K9P (@Roger Olivella). I think that they have limited resources though

Thank you for the useful info Jose. Would a call involving those guys and the Altos guys who are asking about this be the most productive way forwards do you think?

I missed the Altos guys messages, could you point me please to the discussion

Sorry, these were not public discussions, we've just been asked to investigate the best course of action (we have a services agreement that includes community contributions)

OK, I think that a meeting could be the way forward, though to make it as successful as possible it should involve as many people interested in proteomics as possible

I also don't know how we would handle the ongoing activity of the #C02Q3FL29PD channel alongside a parallel development effort. I'm surprise that's still there given that the workflow is archived.

If anyone wants to pick up the quantms pipeline, that's fine. We just fork it back, and as long as the new maintainers want to follow the guidelines and be a part of nf-core, that's fine. I can't remember if we transferred the repo to them or not, but we shouldn't have. It sets a bad precedent if you can just throw up your hands and walk away with an nf-core repo.

Sorry, I'm completely out of the loop here, what has happened?

I'm a bit confused - why can't Altos just use/contribute to QuantMS? It's still alive and kicking, just had a minor release a monthago: https://github.com/bigbio/quantms

I think the current state is confusing. The workflow on the nf-core website is archived, and yet the channel here is open and active. Is it an nf-core workflow or isn't it? Is that going to be something we do now- you get an entry/ support forum in nf-core but get to do your own thing and follow standards when it suits you? But I'm only being an avatar here, my views don't really matter.

no reason why not to have other proteomics pipelines within nf-core

What about the one-pipeline-for-one-thing thing? Surely we can't have another one occupying exactly the same space?

why do the Altos people specifically want something within nf-core

I'm extrapolating slightly, but I think Altos perceive the workflow as no longer part of nf-core, they're a big supporter of the nf-core community, and they'd rather support development within that community framework than within the arbitrary fiefdom of another org

1. I wasn't aware about #C02Q3FL29PD still being open... I can't remember if we agreed on keeping that open as a part of our parting (@Phil Ewels?) 2. Maybe I misunderstood exactly what you meant here. If you're referring specifically to quantms, I would say there is no reason a developer could make a proposal that makes an exact parallel, but probably in the discussion we would strongly encourage them to contribute to a friendly pipeline we trust first (i.e. quantms), before we accept it if the proposer REALLY doesn't like the other pipeline but would encourage them to actually make it distinct... 3. I see... I dunno what we could do about that (see point 3). I'm all for reducing redundency across all of bioinformatics if we can... but I think it's a bit silly to pick a pipeline not based on quality but purely because essentially of 'branding' (as in quantms have their own active community afaik) But these are my own perosnal opinions... thoughts <!subteam^S016E36N7TQ>?

I'm content to be given the 'official line' on this, even if that's "we feel that the best course of action at this time is to continue working with bigbio/quantms"

(I don't know if we have the official line yet, so lets see what everyone else says 🙂 )

Another topic to discuss at the core retreat?

I guess need proper deprecation docs

OK, I'll rely on you all powerful ones to discuss that next week, and seek an official line thereafter

haven't read everything, but:

I can't remember if we transferred the repo to them or not, but we shouldn't have. It sets a bad precedent if you can just throw up your hands and walk away with an nf-core repo.

it was never moved to nf-core but was always a fork. This is from earlier days when we were less stringent on enforcing it

Are you sure? I'm pretty sure it WAS in nf-core, but they jsut developed on the fork all the time