Create subworkflow for deconvolution analyses in bulk data

[apeltzer]

Description of feature

Celltype deconvolution could be a subworkflow in nf-core and then be imported (with some adjustments) into multiple pipelines, one of them being differentialabundance as something users could turn on / off on demand.

This could encompass:

• Create nf-core module for immunedeconv and add to pipeline #368

• omnideconv (?)
• gene set analysis via GSVA (not sure about overlaps with other functionality here already - but can be discussed with community)

x-ref @christopher-mohr same subworkflow would be interesting for e.g. nanostring.