Automatic design of novel potential 3CL$^{\text{pro}}$ and PL$^{\text{pro}}$ inhibitors

TL;DR

This paper introduces MONAS, a framework combining GNNs, energy models, and MCTS to design potential inhibitors for SARS-CoV-1 and SARS-CoV-2, successfully identifying promising molecules from a large search space.

Contribution

It presents a novel integrated molecule optimization framework using neural networks and search algorithms for drug discovery.

Findings

Identified 120,000 candidate molecules likely to inhibit SARS-CoV-1.

Demonstrated the effectiveness of combining GNNs, energy models, and MCTS.

Explored 40 million molecules to find promising candidates.

Abstract

With the goal of designing novel inhibitors for SARS-CoV-1 and SARS-CoV-2, we propose the general molecule optimization framework, Molecular Neural Assay Search (MONAS), consisting of three components: a property predictor which identifies molecules with specific desirable properties, an energy model which approximates the statistical similarity of a given molecule to known training molecules, and a molecule search method. In this work, these components are instantiated with graph neural networks (GNNs), Deep Energy Estimator Networks (DEEN) and Monte Carlo tree search (MCTS), respectively. This implementation is used to identify 120K molecules (out of 40-million explored) which the GNN determined to be likely SARS-CoV-1 inhibitors, and, at the same time, are statistically close to the dataset used to train the GNN.