Evaluation of machine learning models for cytochrome P450 3A4, 2D6, and 2C9 inhibition

In this study, we systematically evaluated the conventional machine learning and deep learning models fo r three major P450 enzymes, CYP3A4, CYP2D6, and CYP2C9 from several perspectives, such as algorithms, molecular representation, and data partitioning strategies. Our results showed that the XGBoost and CatBoost algorithms coupled with the combined fingerprint/physicochemical descriptor features exhi bited the best performance with Area Under Curve (AUC)  of 0.92, while the deep learning models were generally inferior to the conventional machine learning models (average AUC reached 0.89) on the same test sets. We also found that data volume and sampling strategy had a minor effect on model perf ormance. We anticipate that these results are helpful for the selection of molecular representations and machine learning/deep learning algorithms in the P450 model construction and the future model development of P450 inhibition.
Source: Journal of Applied Toxicology - Category: Toxicology Authors: Tags: RESEARCH ARTICLE Source Type: research