Enhanced-Sampling Simulations for the Estimation of Ligand Binding Kinetics: Current Status and Perspective

Abstract

The dissociation rate (k _off) associated with ligand unbinding events from proteins is a parameter of fundamental importance in drug design. Here we review recent major advancements in molecular simulation methodologies for the prediction of k _off. Next, we discuss the impact of the potential energy function models on the accuracy of calculated k _off values. Finally, we provide a perspective from high-performance computing and machine learning which might help improve such predictions.

Keywords: QM/MM; drug discovery; enhanced sampling; kinetics; machine learning; molecular dynamics; parallel computing.

FIGURE 1

Schematic of a LiGaMD Simulation. The LiGaMD potential (∆U_boost) acts when the potential energy of a protein-ligand complex (black line) is below a predefined threshold (dashed line), adding a harmonic potential to raise the energy of the system (cyan line) and favor the exploration of the conformational space of the ligand-protein complex.

FIGURE 2

Schematic of a Metadynamics Simulation. On the CV-projected FES (red line), MetaD deposits a series of gaussians that sum up (from dark blue to white) until the system becomes diffusive in the CV space. This approach can be exploited to reduce the barrier height to have a reasonable transition time and reweight it a posteriori for an estimation of the kinetic constants (see Text).

FIGURE 3

Simplified schematic depiction of the MSM construction pipeline. (A) Several continuous MD trajectories are simulated in parallel. (B) The trajectories are discretized. (C) A reversible transition probability matrix is calculated from a matrix of state-to-state transition counts (D) Probability fluxes between states (gray arrows, with line thickness representing the magnitude of the flux) indicate the highest likelihood transition paths and can be used to calculate the mean first passage time (MFPT) between states.