R with conda in nextflow

[hkaspersen]

For a report process, I am using a docker image, with R, that has been manually created and uploaded online to dockerhub. This R image contains specific R library versions in addition to R itself.
However, I want my pipeline to be able to use conda as well as docker/singularity. My question is then: what is best practices here? Do I have to define a conda environment yml file and refer to it in the process definition, or how is it done?

My processes have the following defined for containers and conda (example from another process than the R process)

	conda (params.enable_conda ? 'bioconda::seqkit=2.1.0' : null)
	container 'quay.io/biocontainers/seqkit:2.1.0--h9ee0642_0'

This works fine for processes with only one program, but I do not know how to define this for a conda env with R + specific package versions.

[jemunro]

In this case it would probably be better to use a conda environment file: https://www.nextflow.io/docs/latest/conda.html#use-conda-environment-files

[hkaspersen]

This only shows how bad I am to read the documentation! Thanks!