Enhancing Molecular Design through Graph-based Topological Reinforcement Learning

TL;DR

This paper introduces GraphTRL, a novel reinforcement learning framework that combines chemical and structural data using advanced graph techniques to enhance molecular generation for drug discovery.

Contribution

It presents GraphTRL, integrating multiscale weighted colored graphs and persistent homology into RL for improved molecular design.

Findings

GraphTRL outperforms existing methods in binding affinity prediction.

The approach effectively incorporates structural information into molecular generation.

Results suggest potential to accelerate drug discovery processes.

Abstract

The generation of drug-like molecules is crucial for drug design. Existing reinforcement learning (RL) methods often overlook structural information. However, feature engineering-based methods usually merely focus on binding affinity prediction without substantial molecular modification. To address this, we present Graph-based Topological Reinforcement Learning (GraphTRL), which integrates both chemical and structural data for improved molecular generation. GraphTRL leverages multiscale weighted colored graphs (MWCG) and persistent homology, combined with molecular fingerprints, as the state space for RL. Evaluations show that GraphTRL outperforms existing methods in binding affinity prediction, offering a promising approach to accelerate drug discovery.