Table 1. Performance of various algorithms to classify nucleotide-binding sites.
DeepDrug3D is compared to volume- and shape-based approaches, a classifier employing the histogram of gradients with principal component analysis (HOG/PCA), pocket matching with G-LoSA, molecular docking with Vina, and sequence signature detection with ScanProsite. The performance is assessed with the accuracy (ACC), precision (PPV), sensitivity (TPR), specificity (TNR), and the area under the curve (AUC).
| Algorithm | ACC | PPV | TPR | TNR | AUC |
|---|---|---|---|---|---|
| DeepDrug3D | 0.943 | 0.951 | 0.896 | 0.971 | 0.986 |
| Pocket volume | 0.783 | - | - | - | 0.512 |
| Pocket shape | 0.739 | 0.654 | 0.730 | 0.746 | 0.824 |
| HOG/PCA | 0.689 | 0.607 | 0.645 | 0.719 | 0.611 |
| ScanProsite | 0.722 | 0.917 | 0.411 | 0.970 | - |
| G-LoSA | 0.730 | 0.750 | 0.589 | 0.843 | 0.770 |
| Vina | 0.634 | 0.578 | 0.648 | 0.622 | 0.701 |