Table 2 Performance comparisons of different methods on D92M over leave-one-out cross-validation. The best results for each metric are in bold
From: Identifying GPCR-drug interaction based on wordbook learning from sequences
Method | Sn (%) | Sp (%) | Acc (%) | Str(%) | MCC |
|---|---|---|---|---|---|
iGPCR-Drug | 78.3 | 91.4 | 86.9 | 84.9 | 0.71 |
OET-KNN | 77.8 | 88.7 | 85.0 | 83.3 | 0.67 |
QuickRBF | 74.8 | 92.4 | 86.4 | 83.6 | 0.69 |
SVM | 74.2 | 92.7 | 86.3 | 83.5 | 0.69 |
RF | 76.5 | 92.9 | 87.3 | 84.7 | 0.71 |
RF + PPP | 79.7 | 92.8 | 88.3 | 86.3 | 0.73 |
Proposed(Base learner) | 81.4 | 84.7 | 83.6 | 83.1 | 0.64 |
Proposed(Ensemble model with Ne = 4) | 81.1 | 87.1 | 85.1 | 84.1 | 0.67 |