Skip to main content

Table 1 Number drugs, targets and positive instances (known interactions) vs. the number of negative (or unknown) interactions on each dataset

From: A multiple kernel learning algorithm for drug-target interaction prediction

  Datasets
  Nuclear receptors GPCR Ion channel Enzyme
Interactions     
Known 90 635 1476 2926
  (6.41 %) (3 %) (3.45 %) (1 %)
Unknown 1314 20550 41364 292554
  (93.59 %) (97 %) (96.55 %) (99 %)
Entity     
Drugs 54 223 210 445
Targets 26 95 204 664