Enhancing Hit Discovery in Virtual Screening through Absolute Protein-Ligand Binding Free-Energy Calculations

In the hit identification stage of drug discovery, adiverse chemicalspace needs to be explored to identify initial hits. Contrary to empiricalscoring functions, absolute protein-ligand binding free-energyperturbation (ABFEP) provides a theoretically more rigorous and accuratedescription of protein-ligand binding thermodynamics and could,in principle, greatly improve the hit rates in virtual screening.In this work, we describe an implementation of an accurate and reliableABFEP method in FEP+. We validated the ABFEP method on eight congenericcompound series binding to eight protein receptors including bothneutral and charged ligands. For ligands with net charges, the alchemicalion approach is adopted to avoid artifacts in electrostatic potentialenergy calculations. The calculated binding free energies correlatewith experimental results with a weighted average of R (2) = 0.55 for the entire dataset. We also observe an overallroot-mean-square error (RMSE) of 1.1 kcal/mol after shifting the zero-pointof the simulation data to match the average experimental values. ThroughABFEP calculations using apo versus holo protein structures, we demonstratedthat the protein conformational and protonation state changes betweenthe apo and holo proteins are the main physical factors contributingto the protein reorganization free energy manifested by the overestimationof raw ABFEP calculated binding free energies using the holo structuresof the proteins. Furthermore, we performed ABFEP calculations in threevirtual screening applications for hit enrichment. ABFEP greatly improvesthe hit rates as compared to docking scores or other methods likemetadynamics. The good performance of ABFEP in rank ordering compoundsdemonstrated in this work confirms it as a useful tool to improvethe hit rates in virtual screening, thus facilitating hit discovery.