Successful drug development requires the attainment of a suitable balance of appropriate pharmacokinetics, sufficient therapeutic activity and acceptable side-effects. Liabilities in any of these areas can cause failure of a project, hence early provision of information on them can prove invaluable in selecting lead series and compounds, thus reducing the likelihood of expensive late stage failure.
As in silico models improve, virtual methods are becoming more widely used, allowing cost effective screening and enabling synthesis to be prioritised on the compounds predicted to have the greatest likelihood of success. Traditionally, virtual models have been used to predict individual properties and rank order the success of a compound based on a specific parameter. However, both drug efficacy and drug-induced toxicity are highly dependent on the pharmacokinetics of a compound, and the concentration of the drug in the blood and tissues. Therefore, it is important to consider pharmacokinetics when assessing predictions of efficacy or toxicity.
Cyprotex have developed a unique virtual prediction tool known as chemPKTM which predicts PK directly from chemical structure using PBPK models optimised directly from human clinical data. chemPKTM can be integrated with tools to predict efficacy (e.g., chemTarget) to provide more clinically relevant predictions by taking exposure into consideration.
chemTarget is the most recent virtual screening tool launched by Cyprotex. The models are built using a unique internally developed pattern recognition system which enables robust models to be generated from binding affinity data for specific drug targets. By combining virtual predictions from chemPKTM and chemTarget, it is possible to predict an indepth analysis of the level of engagement in vivo at the range of concentrations predicted by chemPKTM.
Learn more about our virtual screening tools