Predicting Molecular Ground-State Conformation via Conformation Optimization

2024-10-13

arXiv.org

AI Summary

Predicting molecular ground-state conformations is critical for drug design and other chemical applications; however, traditional energy-optimization methods are computationally expensive, while existing deep learning approaches compromise accuracy and interpretability. To address this, we propose WGFormer—a novel autoencoding architecture that integrates Wasserstein gradient flows with an SE(3)-equivariant Transformer. This work is the first to embed Wasserstein gradient flow theory into an SE(3)-equivariant neural network, enabling implicit learning of physically grounded energy-minimization dynamical trajectories. Coupled with latent-space atomic mixture modeling and an MLP-based decoder, WGFormer achieves efficient, high-accuracy, and interpretable conformational optimization. On multiple benchmark datasets, WGFormer significantly outperforms state-of-the-art methods, achieving lower RMSD errors and stronger correlation with quantum-mechanical energies—demonstrating superior physical consistency and generalization capability.

Technology Category

Machine Learning: Deep Generative Models & AutoencodersSearch and Optimization: Learning to SearchComputer Vision: Learning & Optimization for CV

Application Category

Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphsResponsible Web: Machine-in-the-loop, human agency and autonomySearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Predicting molecular ground-state conformation (i.e., energy-minimized conformation) is crucial for many chemical applications such as molecular docking and property prediction. Classic energy-based simulation is time-consuming when solving this problem while existing learning-based methods have advantages in computational efficiency but sacrifice accuracy and interpretability. In this work, we propose a novel and effective method to bridge the energy-based simulation and the learning-based strategy, which designs and learns a Wasserstein gradient flow-driven SE(3)-Transformer, called WGFormer, for molecular ground-state conformation prediction. Specifically, our method tackles this task within an auto-encoding framework, which encodes low-quality conformations by the proposed WGFormer and decodes corresponding ground-state conformations by an MLP. The architecture of WGFormer corresponds to Wasserstein gradient flows -- it optimizes molecular conformations by minimizing an energy function defined on the latent mixture models of atoms, thereby significantly improving performance and interpretability. Extensive experiments show that our method consistently outperforms state-of-the-art competitors, providing a new and insightful paradigm to predict molecular ground-state conformation.

Problem

• Predict molecular ground-state conformation accurately.
• Bridge energy-based and learning-based methods effectively.
• Enhance computational efficiency and interpretability simultaneously.

Innovation

• Wasserstein gradient flow-driven SE(3)-Transformer
• Auto-encoding framework for conformation prediction
• MLP decodes ground-state conformations