RFE: parallelized CBS (ParDNAcopy)

CenterForMedicalGeneticsGhent / WisecondorX

WisecondorX — An evolved WISECONDOR

Other

94 stars 41 forks source link

RFE: parallelized CBS (ParDNAcopy) #97

Open lbeltrame opened 1 year ago

lbeltrame commented 1 year ago

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 commented 1 year ago

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.