use peppy

[stolarczyk]

@vreuter, @nsheff, @johanneskoester

This is a reiteration of #8, but with a re-written version of peppy that supports the config format revamp discussed in pepkit/peppy#334

[stolarczyk]

sample_table_index and subsample_table_index arguments could be skipped if snakemake followed peppy defaults:

• "sample_name" for sample_table index colnames
• ["sample_name", "subsample_name"] for subsample_table index colnames

[nsheff]

@stolarczyk can you update this to the now-almost-final peppy updates when you get a chance?

[nsheff]

Hi @johanneskoester can you take a look at this when you get a moment?

[nsheff]

FYI we released peppy so it's now live with all the updates we discussed.

[johanneskoester]

Hi Nathan,
yes, I'll have a look. Still not a fan of the snake_config tag, but maybe it can be used in a different way. I just cannot promise that I will be able to evaluate it immediately.

Best,
Johanne

[nsheff]

Still not a fan of the snake_config tag

that's not build in to peppy -- it's just used here. you could instead just put all options in config.yaml directly into the PEP... is that what you're wanting?

[johanneskoester]

Thanks for the clarification! I will probably keep the PEP and the pipeline config separate then. My idea is that the PEP should just describe the samples, while the pipeline describes the analysis. This way, one can easily plug the very same PEP into a completely different pipeline (other WMS, other analysis, ...).