Alchemical Free Energy Methods Applied to Complexes of the First Bromodomain of BRD4

Accurate and rapid predictions of the binding affinity of a compound to atarget are one of the ultimate goals of computer aided drug design. Alchemicalapproaches to free energy estimations follow the path from an initial state of the systemto thefinal state through alchemical changes of the energy function during a moleculardynamics simulation. Herein, we explore the accuracy and efficiency of two suchtechniques: relative free energy perturbation (FEP) and multisite lambda dynamics(MS lambda D). These are applied to a series of inhibitors for the bromodomain-containingprotein 4 (BRD4). We demonstrate a procedure for obtaining accurate relative bindingfree energies using MS lambda D when dealing with a change in the net charge of the ligand.This resulted in an impressive comparison with experiment, with an average difference of0.4 +/- 0.4 kcal mol-1. In a benchmarking study for the relative FEP calculations, we foundthat using 20 lambda windows with 0.5 ns of equilibration and 1 ns of data collection foreach window gave the optimal compromise between accuracy and speed. Overall, relativeFEP and MS lambda D predicted binding free energies with comparable accuracy, an average of0.6 kcal mol-1for each method. However, MS lambda D makes predictions for a larger molecular space over a much shorter time scale thanrelative FEP, with MS lambda D requiring a factor of 18 times less simulation time for the entire molecule space

Automated summary

This study explores the accuracy and efficiency of two alchemical approaches, relative free energy perturbation (FEP) and multisite lambda dynamics (MS λ D), in predicting binding affinity of inhibitors for a specific protein. The results show that both methods have comparable accuracy in predicting binding free energies, with an average difference of 0.6 kcal mol-1. However, MS λ D is able to predict for a larger molecular space in a much shorter time scale compared to relative FEP.