Efficient Molecular Conformer Generation with SO(3)-Averaged Flow Matching and Reflow

2025-07-13

AI Summary

Existing diffusion and flow-based molecular conformation generation methods suffer from high computational overhead, hindering training and sampling efficiency. This paper proposes an efficient 3D conformation generation framework based on flow matching. Our key innovations are: (1) an SO(3)-averaged flow training objective that explicitly encodes rotational symmetry to accelerate convergence; and (2) a reconstruction distillation mechanism that compresses multi-step flow matching into single-step sampling. The method achieves substantial efficiency gains without compromising generation quality: it accelerates training convergence and enables high-fidelity, chemically valid conformation generation in just one inference step. On GEOM-QM9 and GEOM-Drugs benchmarks, our approach attains state-of-the-art performance with significantly reduced training time. This work establishes a scalable, high-throughput paradigm for rapid conformational sampling in drug discovery.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Machine Learning: Learning with Manifolds
• Computer Vision: Diffusion Models for Vision

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines

Abstract

Fast and accurate generation of molecular conformers is desired for downstream computational chemistry and drug discovery tasks. Currently, training and sampling state-of-the-art diffusion or flow-based models for conformer generation require significant computational resources. In this work, we build upon flow-matching and propose two mechanisms for accelerating training and inference of generative models for 3D molecular conformer generation. For fast training, we introduce the SO(3)-Averaged Flow training objective, which leads to faster convergence to better generation quality compared to conditional optimal transport flow or Kabsch-aligned flow. We demonstrate that models trained using SO(3)-Averaged Flow can reach state-of-the-art conformer generation quality. For fast inference, we show that the reflow and distillation methods of flow-based models enable few-steps or even one-step molecular conformer generation with high quality. The training techniques proposed in this work show a path towards highly efficient molecular conformer generation with flow-based models.

Problem

• Accelerate training of 3D molecular conformer generation models
• Improve inference speed for high-quality conformer generation
• Reduce computational resources for flow-based generative models

Innovation

• SO(3)-Averaged Flow for fast training
• Reflow for few-step high-quality inference
• Distillation enables one-step conformer generation