Conformation Generation using Transformer Flows

TL;DR

ConfFlow is a novel transformer-based flow model that efficiently generates accurate 3D molecular conformations directly in coordinate space, outperforming existing methods especially for large molecules.

Contribution

Introduces ConfFlow, a flow-based transformer model that directly samples molecular conformations without explicit physical constraints, improving scalability and accuracy.

Findings

ConfFlow improves conformation accuracy by up to 40% for large molecules.

The model enables fast and interpretable conformation generation.

It outperforms state-of-the-art methods in molecular modeling tasks.

Abstract

Estimating three-dimensional conformations of a molecular graph allows insight into the molecule's biological and chemical functions. Fast generation of valid conformations is thus central to molecular modeling. Recent advances in graph-based deep networks have accelerated conformation generation from hours to seconds. However, current network architectures do not scale well to large molecules. Here we present ConfFlow, a flow-based model for conformation generation based on transformer networks. In contrast with existing approaches, ConfFlow directly samples in the coordinate space without enforcing any explicit physical constraints. The generative procedure is highly interpretable and is akin to force field updates in molecular dynamics simulation. When applied to the generation of large molecule conformations, ConfFlow improve accuracy by up to $40\%$ relative to state-of-the-art learning-based methods. The source code is made available at https://github.com/IntelLabs/ConfFlow.