Optimizing Molecules using Efficient Queries from Property Evaluations

TL;DR

QMO is a query-based molecule optimization framework that leverages latent embeddings to efficiently improve molecular properties, outperforming existing methods in benchmark and real-world drug discovery tasks.

Contribution

Introduction of QMO, a novel query-based framework utilizing autoencoder embeddings for efficient molecule optimization with property constraints.

Findings

QMO outperforms existing methods in drug-likeness and solubility optimization.

QMO effectively improves SARS-CoV-2 inhibitor affinity.

QMO reduces toxicity in antimicrobial peptides.

Abstract

Machine learning based methods have shown potential for optimizing existing molecules with more desirable properties, a critical step towards accelerating new chemical discovery. Here we propose QMO, a generic query-based molecule optimization framework that exploits latent embeddings from a molecule autoencoder. QMO improves the desired properties of an input molecule based on efficient queries, guided by a set of molecular property predictions and evaluation metrics. We show that QMO outperforms existing methods in the benchmark tasks of optimizing small organic molecules for drug-likeness and solubility under similarity constraints. We also demonstrate significant property improvement using QMO on two new and challenging tasks that are also important in real-world discovery problems: (i) optimizing existing potential SARS-CoV-2 Main Protease inhibitors toward higher binding affinity; and (ii) improving known antimicrobial peptides towards lower toxicity. Results from QMO show high consistency with external validations, suggesting effective means to facilitate material optimization problems with design constraints.