Machine-Learning Model for Predicting the Rate Constant of ProteinLigand Dissociation

Abstract

An increasing number of recent studies have shown that the binding kinetics of a drug molecule to its target correlates strongly with its efficacy in vivo. Therefore, ligand optimization oriented to improved binding kinetics provides new ideas for rational drug design. Currently, ligand binding kinetics is modeled mainly through extensive molecular dynamics simulations, which limits its application to real-world problems. The present study aimed at obtaining a general-purpose quantitative structure-kinetics relationship (QSKR) model for predicting the dissociation rate constant (koff) of a ligand based on its complex structure. This type of model is expected to be suitable for highthroughput tasks in structure-based drug design. We collected the experimentally measured koff values for 406 ligand molecules from literature, and then constructed a three-dimensional structural model for each protein-ligand complex through molecular modeling. A training set was compiled using 60% of those complexes while the remaining 40% were assigned to two test sets. Based on distance-dependent protein-ligand atom pair descriptors, a random forest algorithm was adopted to derive a QSKR model. Various random forest models were then generated based on the descriptor sets obtained under different conditions, such as distance cutoff, bin width, and feature selection criteria. The cross-validation results of those models were then examined. It was observed that the optimal model was obtained when the distance cutoff was 15 Å (1 Å = 0.1 nm), the bin width was 3 Å, and feature selection variance level was 2. The final QSKR model produced correlation coefficients around 0.62 on the two independent test sets. This level of accuracy is at least comparable to that of the predictive models described in literature, which are typically computationally much more expensive. Our study attempts to address the issue of predicting koff values in drug design. We hope that it can provide inspiration for further studies by other researchers.

Baseline Model for Predicting Protein–Ligand Unbinding Kinetics through Machine Learning

“…We report that selected RF-Score-based descriptors for HSP90 and p38 MAP kinase subsets perform better than the reference models based on molecular descriptors, reported previously to correlate with binding kinetics and SIFP fingerprints. Similarly, Su et al reported a higher correlation for extended RF-Score descriptors over molecular descriptors for the HSP90 set, starting with robust dataset preparation. The research in the direction of high-throughput methods for binding kinetics is currently being continued.…”

“…We assume that the significant difference in the test set results is partly due to the fact that our models were tested on external test sets with various protein structures. It should be noted that a highly similar random forest approach had recently been applied in Su’s work for binding kinetics predictions. The authors curated a database composed of 406 protein–ligand complexes with k off data and constructed ML models based on the extended RF-score descriptors.…”

“…work60 for binding kinetics predictions. The authors curated a database composed of 406 protein−ligand complexes with k off data and constructed ML models based on the extended RFscore descriptors.…”

Predicting Rate Constants of Hydroxyl Radical Reactions with Alkanes Using Machine Learning

“…17 Su et al adopted distance-dependent protein−ligand atom pair descriptors and a random forest algorithm to derive a quantitative structure−kinetics relationship model for predicting the dissociation rate constant of a ligand. 18 In this work, neural network (NN) models are developed to predict the temperature-dependent rate constants of the title reactions with the molecular structure descriptors as input. Particular attention is paid to predicting the site-specific rate constants.…”

“…Chen and Goldsmith utilized a machine learning surrogate potential energy surface to replace the direct CASPT2 energy in variational transition state calculations, by which the computational costs were greatly reduced . Su et al adopted distance-dependent protein–ligand atom pair descriptors and a random forest algorithm to derive a quantitative structure–kinetics relationship model for predicting the dissociation rate constant of a ligand …”