RFE: parallelized CBS (ParDNAcopy)

[lbeltrame]

A possible improvement in segmentation speed would be to use ParDNAcopy to parallelize the actual CBS.

Pros:

• Gets things done quicker (always a plus)
• Setting a random seed (in ParDNAcopy::parSegment) allows more reproducible resutls

Cons:

• One more dependency
• No more updates since 2014 (it's not like DNAcopy got much love either, though)

Again, if this is deemed worthwhile I'm willing to actually implement that.

[matthdsm]

Hi Luca,

Away on holidays atm, but we welcome each improvement you can throw at us!
I'm finishing up some other stuff over at nf-core and then I'll re-focus on wisecondorx for a while.
I've got some code improvements in the pipeline and it'd be nice to have some kind of CI here so we don't regress while adding new functionality.