Navigating the Design Space of Equivariant Diffusion-Based Generative Models for De Novo 3D Molecule Generation

TL;DR

This paper introduces EQGAT-diff, a novel E(3)-equivariant diffusion model for 3D molecule generation that outperforms existing models in accuracy, efficiency, and transferability, advancing de novo molecular design.

Contribution

The paper presents EQGAT-diff, a new diffusion model that incorporates continuous atom positions and categorical chemical features, with improved training convergence and sample quality.

Findings

EQGAT-diff outperforms established models on QM9 and GEOM-Drugs datasets.

Including chemically motivated features enhances molecule validity.

Fine-tuning on limited data improves performance and transferability.

Abstract

Deep generative diffusion models are a promising avenue for 3D de novo molecular design in materials science and drug discovery. However, their utility is still limited by suboptimal performance on large molecular structures and limited training data. To address this gap, we explore the design space of E(3)-equivariant diffusion models, focusing on previously unexplored areas. Our extensive comparative analysis evaluates the interplay between continuous and discrete state spaces. From this investigation, we present the EQGAT-diff model, which consistently outperforms established models for the QM9 and GEOM-Drugs datasets. Significantly, EQGAT-diff takes continuous atom positions, while chemical elements and bond types are categorical and uses time-dependent loss weighting, substantially increasing training convergence, the quality of generated samples, and inference time. We also showcase that including chemically motivated additional features like hybridization states in the diffusion process enhances the validity of generated molecules. To further strengthen the applicability of diffusion models to limited training data, we investigate the transferability of EQGAT-diff trained on the large PubChem3D dataset with implicit hydrogen atoms to target different data distributions. Fine-tuning EQGAT-diff for just a few iterations shows an efficient distribution shift, further improving performance throughout data sets. Finally, we test our model on the Crossdocked data set for structure-based de novo ligand generation, underlining the importance of our findings showing state-of-the-art performance on Vina docking scores.