Manas02/ChemEncoder

Issue with Data curation

Closed this issue · 2 comments

One of the paper that we are using as benchmark is this
The Dataset section is not very helpful (gives reference to obscure papers and no links or code is provided in support)

O, All Wise! Help me !!!

@Manas02 I am mostly very sceptical about the Chinese authors.
Nevertheless, here are some links to curate the dataset

  1. https://zinc15.docking.org/trials/
  2. https://zinc15.docking.org/substances/subsets/world/ (This may be a total of points 3 and 4)
  3. https://zinc15.docking.org/substances/subsets/world-not-fda/
  4. https://zinc15.docking.org/substances/subsets/fda/
  5. https://zinc15.docking.org/substances/subsets/in-man-only/
  6. https://zinc15.docking.org/substances/subsets/aggregators/ ( This is a special category that leads to false-positive hits, label them as non-drug like)
  7. https://zinc15.docking.org/substances/subsets/standard-ok/ (This as external needs to be classified to show how applicable our method is.)

I hope this should help

Regards
Elvis

Thanks boss