Metabolism plays a significant role in failure of drugs to carry out their intended action in the body. Drugs and other xenobiotics in the body are mainly biotransformed through oxidation reactions catalyzed by cytochrome P450 enzymes, in particular the CYP3A4 isoform. This enzyme is well known for its very wide substrate specificity and promiscuity. Understanding where in a compound's structure the enzymatic oxidation would take place, i.e. site of metabolism (SOM), is crucial for successful modification of molecules in order to avoid unwanted metabolism. Several in silico methods have been developed for SOM predictions including docking based methods. One of the benefits of docking based methods is that they provide insight into the structure of the drug-enzyme complex. However, crystallographic studies of CYP3A4 have showed that the structure of this enzyme may obtain substantially different conformations depending on the structure of the ligand it is cocrystallized with. Therefore, it is not straight forward to decide which of the different crystal structures of CYP3A4 should be used for SOM prediction when using docking methods. Here we show that simple discriminative models allow selection of CYP3A4 structural conformer to yield better SOM predictions for each individual substrate.
[1]
Lars Carlsson,et al.
State-of-the-art Tools for Computational Site of Metabolism Predictions: Comparative Analysis, Mechanistical Insights, and Future Applications
,
2007,
Drug metabolism reviews.
[2]
Jose Cosme,et al.
Crystal Structures of Human Cytochrome P450 3A4 Bound to Metyrapone and Progesterone
,
2004,
Science.
[3]
T. Sjögren,et al.
Structural basis for ligand promiscuity in cytochrome P450 3A4
,
2006,
Proceedings of the National Academy of Sciences.
[4]
H. van de Waterbeemd,et al.
ADMET in silico modelling: towards prediction paradise?
,
2003,
Nature reviews. Drug discovery.
[5]
Doo Nam Kim,et al.
New Combined Model for the Prediction of Regioselectivity in Cytochrome P450/3A4 Mediated Metabolism
,
2008,
J. Chem. Inf. Model..