Combining selectivity and affinity predictions using an integrated support vector machine (SVM) approach: a novel tool to discriminate between the human adenosine A2A and A3 receptor pyrazolo-triazolo-pyrimidine antagonists binding sites

A large number of drugs have efficacy problems and limiting side effects because the compounds do not differentiate between receptor subtypes. There is thus considerable interest in attaining therapeutic selectivity in order to obtain drugs without, or at least with less, side effects while retaining the desired function. Due to the potential role of A2A and A3 adenosine receptors (AR) in several physiopathological processes, such as neurodegenerative diseases for A2A ARs and tumor growth and glaucoma for A3 ARs, it is our goal to design compounds showing high selectivity between these two receptor subtypes [1]. Nowadays, very few valuable computational tools are available for the prediction of receptor subtypes selectivity. In the present study, we present an alternative application of the support vector machine (SVM) and support vector regression (SVR) methodologies to simultaneously describe both A2A AR versus A3 AR subtype selectivity profiles and the corresponding receptor binding affinities. We have implemented an integrated application of SVM–SVR approach based on the use of our recently reported autocorrelated molecular descriptors encoding for the molecular electrostatic potential (autoMEP) to simultaneously discriminate A2A AR versus A3 AR antagonists and to predict their binding affinity to the corresponding receptor subtype of a large dataset of known pyrazolo-triazolo-pyrimidine analogs [2–5]. To validate our approach, we have synthesized 51 new pyrazolo-triazolo-pyrimidine derivatives. Predictions for both A2A AR/A3 AR subtype selectivity and receptor binding affinity profiles are very encouraging.