Table 5.
Summary of the 3-D PhypI/3-D QSAR model ensemble experimental/structure-based/ligand-based predictive ability for TSCRY.
| Entry | pKi | EC Pred. pKi a | AAEP d | SB Pred. pKi a | AAEP d | LB Pred. pKi a | AAEP d | |||
|---|---|---|---|---|---|---|---|---|---|---|
| LOO b | LSO c | LOO b | LSO c | LOO b | LSO c | |||||
| 3ERT | 9.60 | 8.76 | 8.64 | 0.90 | 8.36 | 8.34 | 1.25 | 7.99 | 8.12 | 1.55 |
| 3UU7 | 8.79 | 8.14 | 6.91 | 1.27 | 8.09 | 7.22 | 1.14 | 7.85 | 7.14 | 1.30 |
| 3UUA | 8.79 | 8.15 | 7.54 | 0.94 | 7.05 | 7.12 | 1.71 | 8.07 | 7.37 | 1.07 |
| 3UUC | 5.70 | 4.36 | 4.39 | 1.33 | 4.45 | 4.06 | 1.45 | 5.67 | 6.77 | 0.55 |
| 4DMA | 5.60 | 6.54 | 7.69 | 1.52 | 7.91 | 7.59 | 2.15 | 8.86 | 7.7 | 2.68 |
| 4MG6 | 6.00 | 4.76 | 4.77 | 1.24 | 4.17 | 3.03 | 2.40 | 4.16 | 4.82 | 1.51 |
| 4MG8 | 10.00 | 8.86 | 8.87 | 1.14 | 9.16 | 7.76 | 1.54 | 8.99 | 8.85 | 1.08 |
| 4MG9 | 6.00 | 7.12 | 6.52 | 0.82 | 6.19 | 4.10 | 1.05 | 4.51 | 5.96 | 0.77 |
| 4MGA | 6.00 | 8.13 | 6.99 | 1.56 | 7.13 | 6.89 | 1.01 | 7.41 | 4.98 | 1.22 |
| 4MGC | 7.00 | 8.66 | 6.7 | 0.98 | 6.36 | 6.54 | 0.55 | 7.58 | 5.85 | 0.87 |
| 4MGD | 6.00 | 7.66 | 9.04 | 2.35 | 8.46 | 7.13 | 1.80 | 9.19 | 9.48 | 3.34 |
| 4TUZ | 10.00 | 8.64 | 8.88 | 1.24 | 9.17 | 7.52 | 1.66 | 9.06 | 8.7 | 1.12 |
| 4ZN9 | 9.60 | 8.96 | 8.92 | 0.66 | 8.74 | 7.06 | 1.70 | 8.78 | 8.49 | 0.97 |
a Predictions were obtained with a 3-D PhypI/3-D QSAR model ensemble optimized with LOO and LSO cross-validations. b Leave-one-out cross-validation. c Leave-some-out cross-validation with 5-random-groups-out. d AAEP, the average absolute error of prediction of LOO and LSO cross-validations.