Skip to main content

Table 1 Performance of \(\Delta\)-AEScore compared to the \(\Delta _\text {vina}\text {RF}\) for affinity prediction on the CASF-2013 and CASF-2016 benchmarks. For \(\Delta\)-AEScore the “hard overlap” between the training and both test sets is removed while for \(\Delta _\text {vina}\text {RF}\) only the “hard overlap” between the training set and CASF-2013 is removed [49, 67]. The best performance for each test set is underlined. RMSE values are given in pK units

From: Learning protein-ligand binding affinity with atomic environment vectors

Model Training set Test set RMSE Pearson’s r
\(\Delta\)-AEScore\(^{\dag }\) Refined 2013 CASF-2013 1.53 0.74
\(\Delta\)-AEScore\(^{\dag }\) (no H) Refined 2013 CASF-2013 1.52 0.74
\(\Delta _\text {vina}\text {RF}\) [49] Refined 2013 CASF 2013 0.69
Vina (optim) CASF-2013 1.82 0.61
\(\Delta\)-AEScore\(^{\dag }\) Refined 2016 CASF-2016 1.34 0.79
\(\Delta\)-AEScore\(^{\dag }\) (no H) Refined 2016 CASF-2016 1.32 0.80
\(\Delta _\text {vina}\text {RF}\) [40, 49] Refined 2013 CASF 2016 0.81
Vina (optim) CASF-2016 1.75 0.59
  1. \(^\dag\) This work