Improving scoring-docking-screening powers of protein–ligand scoring functions using random forest

The development of new protein–ligand scoring functions using machine learning algorithms, such as random forest, has been of significant interest. By efficiently utilizing expanded feature sets and a large set of experimental data, random forest based scoring functions (RFbScore) can achieve better correlations to experimental protein–ligand binding data with known crystal structures; however, more extensive tests indicate that such enhancement in scoring power comes with significant under-performance in docking and screening power tests compared to traditional scoring functions. In this work, to improve scoring-docking-screening powers of protein–ligand docking functions simultaneously, we have introduced a Δ_vinaRF parameterization and feature selection framework based on random forest. Our developed scoring function Δ_vinaRF₂₀, which employs 20 descriptors in addition to the AutoDock Vina score, can achieve superior performance in all power tests of both CASF-2013 and CASF-2007 benchmarks compared to classical scoring functions. The Δ_vinaRF₂₀ scoring function and its code are freely available on the web at: https://www.nyu.edu/projects/yzhang/DeltaVina.