Enhancing the effectiveness of virtual screening by using the ChemBioServer: Application to the discovery of PI3Kα inhibitors

The application of an in-house developed web server called ChemBioServer to the filtering and selection of drug candidates for the inhibition of the PI3Kα protein is presented. 1000 candidate molecules were initially selected from a virtual screening experiment. Those molecules were then filtered for steric clashes, physicochemical and toxicity properties and grouped into clusters using the ChemBioServer web application. During this filtering process, 400 compounds were rejected and the remaining 600 were clustered in 20 different groups, allowing for a more efficient visual inspection of the compounds. Representatives of these clusters were then selected for further experimental study. Four out of the seven selected molecules inhibited PI3Kα activity in vitro, indicating that the workflow described herein can be successfully applied in drug discovery. ChemBioServer proved to assist the post-processing application of top-ranked molecules resulting from a docking exercise by increasing the efficiency and the quality of compound selection that passed to the experimental test phase.