search

robinweide / GENOVA

GENome Organisation Visual Analytics
GNU General Public License v3.0
69 stars 15 forks source link

set up GENOVAresult object #108

Closed teunbrand closed 4 years ago

teunbrand commented 5 years ago

Codename: Discovery

teunbrand commented 5 years ago

For the following:

teunbrand commented 5 years ago

Group by sample or by result type?

Below are conceptual options.

Option 1:

discovery <- structure(list(
  sampleA = list(
    result1, # e.g. observed over expected
    result2  # e.g. background or signal
  ),
  sampleB = list(
    result1,
    result2
  )
), class = "assayX_discovery", "discovery", "list")

Option 2:

discovery <- structure(list(
  result1 = array(
    sampleA, 
    sampleB
  ),
  result2 = array(
    sampleA, 
    sampleB
  )
), class = "assayX_discovery", "discovery", "list")

Option 1 is easy to merge the results of different samples, option 2 is easy for computation and visualisation. @robinweide do you have strong opinions?

robinweide commented 5 years ago

I would go for option 2. This would also allow for easy substitution of results by the user afterwards (e.g. pushing the same expected matrices for multiple samples).

robinweide commented 5 years ago

For the following:

  • [ ] APA
  • [ ] ATA
  • [ ] PESCAn
  • [ ] RCP
  • [ ] Saddle
  • [ ] inter/intra TAD contacts
  • [ ] domainogram
teunbrand commented 4 years ago

All these functions now have associated discovery classes.