Skip to main content

Table 7 Results on the protein-protein docking benchmark 2.0 dataset.

From: Exploiting residue-level and profile-level interface propensities for usage in binding sites prediction of proteins

Subset No. of Protein Methoda Precision Recall F1 Accuracy CC
Rigid body 63 AA 0.393 0.447 0.418 0.848 0.301
   BP 0.446 0.495 0.469 0.857 0.328
Medium difficult 13 AA 0.356 0.405 0.379 0.810 0.258
   BP 0.412 0.464 0.436 0.821 0.271
Difficult 8 AA 0.362 0.384 0.372 0.813 0.299
   BP 0.409 0.427 0.428 0.819 0.317
All 84 AA 0.370 0.412 0.390 0.824 0.286
   BP 0.422 0.462 0.441 0.832 0.305
  1. aAA, the classifiers with residue interface propensities as extra features; BP, the classifiers with binary profile interface propensities as extra features.