I've been using the #CFP55NZ5G pipeline and noticed that several conda environments are broken for various reasons. For example, the deeptools modules use deeptools pinned to version 3.5.5 (example here), which is not compatible with NumPy version 2, but the bioconda recipe (meta.yaml) for deeptools version 3.5.5 only specifies a lower limit on NumPy versions (

numpy >=1.9.0

), which inadvertently allows conda to install NumPy version 2 and deeptools version 3.5.5 in the same environment, leading to incompatibilities and pipeline errors. The most straightforward solution is probably just to pin deeptools to version 3.5.6, which enables NumPy v2 compatibility. Since deeptools is an nf-core module (and not a local module to the nf-core/chipseq pipeline), where should I submit such issues and pull requests - to https://github.com/nf-core/chipseq or to https://github.com/nf-core/modules?

Pull request for module should be in modules, to fix the module for everyone, and then the next step would be a PR in chipseq, updating the module in that pipeline

Got it. Is there a standard or automatic way by which updates to modules in

nf-core/modules

are propagated to workflows that use those modules, or would I need to manually copy the changes from

nf-core/modules

to

nf-core/chipseq

?

You can use the nf-core tools to update the nfcore modules in a pipeline https://nf-co.re/docs/nf-core-tools/modules/update

Cool, thanks!