Applications of Modular Co-Design for De Novo 3D Molecule Generation

TL;DR

Megalodon, a scalable transformer model with equivariant layers, advances 3D molecule generation by achieving state-of-the-art results in validity, realism, and energetics, significantly outperforming prior models.

Contribution

Introduction of Megalodon, a scalable transformer with equivariant layers, trained with a joint denoising objective, setting new benchmarks in 3D molecule generation.

Findings

Megalodon achieves state-of-the-art results in 3D molecule generation.

Doubling parameters to 40M improves validity and energy levels.

Megalodon generates up to 49x more valid large molecules.

Abstract

De novo 3D molecule generation is a pivotal task in drug discovery. However, many recent geometric generative models struggle to produce high-quality 3D structures, even if they maintain 2D validity and topological stability. To tackle this issue and enhance the learning of effective molecular generation dynamics, we present Megalodon-a family of scalable transformer models. These models are enhanced with basic equivariant layers and trained using a joint continuous and discrete denoising co-design objective. We assess Megalodon's performance on established molecule generation benchmarks and introduce new 3D structure benchmarks that evaluate a model's capability to generate realistic molecular structures, particularly focusing on energetics. We show that Megalodon achieves state-of-the-art results in 3D molecule generation, conditional structure generation, and structure energy benchmarks using diffusion and flow matching. Furthermore, doubling the number of parameters in Megalodon to 40M significantly enhances its performance, generating up to 49x more valid large molecules and achieving energy levels that are 2-10x lower than those of the best prior generative models.