There are many tools for predicting individual ADME properties from chemical structures using QSAR-based approaches. Predicting pharmacokinetics directly from chemical structure is potentially more challenging, yet ultimately of greater value, as it obviates the need to appropriately extrapolate from a predicted ADME property to its ultimate effect on pharmacokinetics in vivo. Cyprotex provides a unique solution to this challenge, using a physiologically-based pharmacokinetic (PBPK) model, parameterised by means of direct optimisation on human clinical pharmacokinetic data.
The first version of chemPKTM was launched in July 2015. Since that time, the product has been further developed to improve functionality. chemPKTM version 2 can now predict both oral and intravenous human pharmacokinetics and enable simulation of repeat-dose pharmacokinetics. As well as predicting a wide range of pharmacokinetic parameters, it can provide valuable information on hepatic metabolism in addition to renal and faecal elimination routes. The performance of chemPKTM in the prediction of key pharmacokinetic parameters following intravenous or oral administration has been demonstrated on an independent set of test compounds.
chemPKTM runs on the freely available, and flexible, KNIME Analytics Platform which is easy to implement and to use. KNIME facilitates integration with other cheminformatics, analytics and modelling tools.
Prediction of Volume of Distribution from Chemical Structure for a Test Set of 62 Compounds