We have prepared the training sets for the five isoforms of CYP P450 that metabolize the majority of xenobiotics: 3A4, 2C9, 2C19, 2D6 and 1A2. Also in the training set were included the reaction of glucoronidation, catalyzed by UGT. A leave-one-out cross-validation (LOO CV) procedure was performed for each of the training sets to validate the quality of the prediction.
Enzyme | Number of Substrates | Positive examples count | Negative examples count | IAP LOO |
CYP3A4 | 960 | 1410 | 18621 | 0.89 |
CYP2D6 | 588 | 726 | 8366 | 0.92 |
CYP2C9 | 446 | 600 | 7471 | 0.92 |
CYP2C19 | 388 | 517 | 6329 | 0.93 |
CYP1A2 | 573 | 888 | 9375 | 0.92 |
UGT | 592 | 906 | 1136 | 0.98 |
Standard | 1109 | 1701 | 16425 | 0.89 |