Machine Learning Analysis of τRAMD Trajectories to Decipher Molecular Determinants of Drug-Target Residence Times

Abstract

Drug-target residence times can impact drug efficacy and safety, and are therefore increasingly being considered during lead optimization. For this purpose, computational methods to predict residence times, τ, for drug-like compounds and to derive structure-kinetic relationships are desirable. A challenge for approaches based on molecular dynamics (MD) simulation is the fact that drug residence times are typically orders of magnitude longer than computationally feasible simulation times. Therefore, enhanced sampling methods are required. We recently reported one such approach: the τRAMD procedure for estimating relative residence times by performing a large number of random acceleration MD (RAMD) simulations in which ligand dissociation occurs in times of about a nanosecond due to the application of an additional randomly oriented force to the ligand. The length of the RAMD simulations is used to deduce τ. The RAMD simulations also provide information on ligand egress pathways and dissociation mechanisms. Here, we describe a machine learning approach to systematically analyze protein-ligand binding contacts in the RAMD trajectories in order to derive regression models for estimating τ and to decipher the molecular features leading to longer τ values. We demonstrate that the regression models built on the protein-ligand interaction fingerprints of the dissociation trajectories result in robust estimates of τ for a set of 94 drug-like inhibitors of heat shock protein 90 (HSP90), even for the compounds for which the length of the RAMD trajectories does not provide a good estimation of τ. Thus, we find that machine learning helps to overcome inaccuracies in the modeling of protein-ligand complexes due to incomplete sampling or force field deficiencies. Moreover, the approach facilitates the identification of features important for residence time. In particular, we observed that interactions of the ligand with the sidechain of F138, which is located on the border between the ATP binding pocket and a hydrophobic transient sub-pocket, play a key role in slowing compound dissociation. We expect that the combination of the τRAMD simulation procedure with machine learning analysis will be generally applicable as an aid to target-based lead optimization.

Molecular Dynamics Simulations in Drug Discovery and Pharmaceutical Development

“…τRAMD was used to predict relative residence times for 70 HSP90 inhibitors (R 2 of 0.86 after removal of outliers) [266]. Analysis of inhibitor-protein contacts in the trajectories using machine learning improved residence time estimates and detected that interactions between a halogen or a methyl group in the loop-binder compounds with Phe138 lead to a longer residence time [269]. τRAMD was also recently applied to estimate relative residence times of ligands dissociating from different mutants of T4 lysozyme [270].…”

In Silico Prediction of the Dissociation Rate Constants of Small Chemical Ligands by 3D-Grid-Based VolSurf Method

“…In additions, the prediction method was further tested by 94 HSP90 inhibitors. The results showed that the predicted R 2 of the 80 inhibitors was 0.75 with MAPE of 0.39 [11].…”

Ligand binding free energy and kinetics calculation in 2020

“…It is interesting to note that the pmf obtained by these simulations allowed a semi‐quantitative estimate of the energy barriers encountered by the ligands during their unbinding that well explains the different experimental ligand unbinding rates. Processing the RAMD trajectory and simulation unbinding times with a machine learning approach, Kokh et al were able to improve the correlation of the in‐silico ligand residence time with the experimental values for 70 and 94 inhibitors of heat shock protein 90 (HSP90) …”