Characteristics of the training sets and prediction accuracy of consensus QSAR models for CYPs inhibitors.
CYPs | Number of compounds | Interval of values | Mean | R2 | Q2 | SD |
---|---|---|---|---|---|---|
1A2 | 1216 | [1.6:8.8] | 5.47 | 0.999 | 0.680 | 0.625 |
2C9 | 1657 | [1.6:9.0] | 5.28 | 0.994 | 0.609 | 0.512 |
2C19 | 930 | [2.8:8.3] | 5.16 | 0.992 | 0.501 | 0.519 |
2D6 | 1588 | [1.2:9.2] | 5.32 | 0.972 | 0.566 | 0.567 |
3A4 | 3299 | [1.4:10.3] | 5.39 | 0.992 | 0.640 | 0.528 |
The results of SAR prediction for classification .
CYPs | Inhibition | Induction | |||
---|---|---|---|---|---|
Npos | IAP, LOO CV | Npos | IAP, LOO CV | ||
1A2 | 3882 | 0.923 | 26 | 0.908 | |
2D6 | 2957 | 0.891 | |||
2C9 | 3673 | 0.855 | 28 | 0.831 | |
2C19 | 3597 | 0.855 | 8 | 0.834 | |
3A4 | 6804 | 0.873 | 78 | 0.873 |