Merge computeMsigOverlap and computeMsigNetwork

[ahmohamed]

Currently, having a list of significant gene sets, we compute the network using the following code:

gs_ovlap = computeMsigOverlap(siggs, thresh = 0.25)
gs_ovnet = computeMsigNetwork(gs_ovlap, msigdb)

I think we can't merge both into one function, with the signature

computeMsigNetwork <- function(siggs, thresh = 0.25) {
  gs_ovlap = computeMsigOverlap(siggs, thresh = 0.25)
  
  # This is the current computeMsigNetwork function
  # We reuse siggs, which is already a msigGsc object
  .computeMsigNetwork(gs_ovlap, siggs)
}

which will should make the code cleaner, and more black-boxed (if that's desirable).

[bhuvad]

The reason I left this separated was to enable independent analysis of the overlap data if users wish to do so. For instance, they can compute all pairwise overlaps (by setting thresh = 0) and then trying out a few filtering approaches manually. This would provide added flexibility without having to recompute overlaps every single time.