Assessing the Performance of Metadynamics and Path Variables in Predicting the Binding Free Energies of p38 Inhibitors

The accurate yet efficient evaluation of the free energy profiles of ligand-target association is a long sought goal in rational drug design. Methods that calculate the free energy along realistic association pathways, such as metadynamics, have been shown to provide reliable profiles, while accounting properly for solvation and target flexibility. However, these approaches usually require prohibitive computational resources and expert human intervention. Here, we show how multiple walkers metadynamics, when performed with optimal path collective variables (PCV), provides in a predetermined amount of computer time an accurate set of free energy profiles for a series of p38 inhibitors. The chosen test set, spanning a wide range of activity, is a challenging benchmark, both for computational methods and for human intuition, as the correct order for the binding affinity cannot be easily guessed. An excellent ranking of the ligands was obtained with minimal human assistance, an important step toward a fully automated pharmaceutical work-flow.