A 3D-QSAR-Driven Approach to Binding Mode and Affinity Prediction

A method for predicting the binding mode of a series of ligands is proposed. The procedure relies on three-dimensional quantitative structure activity relationships (3D-QSAR) and does not require structural knowledge of the binding site. Candidate alignments are automatically built and ranked according to a consensus scoring function. 3D-QSAR analysis based on the selected binding mode enables affinity prediction of new drug candidates having less than 10 rotatable bonds.