Structure-aware generation of drug-like molecules

TL;DR

This paper introduces a structure-aware generative model for drug-like molecules that incorporates 3D pocket information, improving binding affinity and drug-likeness predictions over baseline methods.

Contribution

A novel supervised model that generates molecules with 3D pose information, integrating structural data into the de-novo design process.

Findings

Improves predicted binding affinities by 8%

Enhances drug-likeness scores by 10%

Proposes molecules with superior binding scores than some known ligands

Abstract

Structure-based drug design involves finding ligand molecules that exhibit structural and chemical complementarity to protein pockets. Deep generative methods have shown promise in proposing novel molecules from scratch (de-novo design), avoiding exhaustive virtual screening of chemical space. Most generative de-novo models fail to incorporate detailed ligand-protein interactions and 3D pocket structures. We propose a novel supervised model that generates molecular graphs jointly with 3D pose in a discretised molecular space. Molecules are built atom-by-atom inside pockets, guided by structural information from crystallographic data. We evaluate our model using a docking benchmark and find that guided generation improves predicted binding affinities by 8% and drug-likeness scores by 10% over the baseline. Furthermore, our model proposes molecules with binding scores exceeding some known ligands, which could be useful in future wet-lab studies.