Scalable Machine Learning Force Fields for Macromolecular Systems Through Long-Range Aware Message Passing

2026-01-07

AI Summary

This work addresses the limitations of existing machine learning force fields, which typically rely on fixed-cutoff local architectures and struggle to accurately capture long-range non-covalent interactions in large molecular systems, leading to error accumulation with system size. To overcome this, the authors construct MolLR25, a high-accuracy DFT reference dataset comprising systems with up to 1,200 atoms, and introduce E2Former-LSR—a novel equivariant Transformer architecture that integrates an explicit long-range attention mechanism. This approach significantly enhances the modeling of complex protein conformations and long-range interactions while maintaining computational efficiency. The model achieves state-of-the-art accuracy on large molecular systems with stable error scaling and demonstrates up to a 30% speedup compared to purely local models.

Technology Category

• Machine Learning: Deep Neural Architectures and Foundation Models
• Computer Vision: Large Vision Models
• Natural Language Processing: (Large) Language Models

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work

Abstract

Machine learning force fields (MLFFs) have revolutionized molecular simulations by providing quantum mechanical accuracy at the speed of molecular mechanical computations. However, a fundamental reliance of these models on fixed-cutoff architectures limits their applicability to macromolecular systems where long-range interactions dominate. We demonstrate that this locality constraint causes force prediction errors to scale monotonically with system size, revealing a critical architectural bottleneck. To overcome this, we establish the systematically designed MolLR25 ({Mol}ecules with {L}ong-{R}ange effect) benchmark up to 1200 atoms, generated using high-fidelity DFT, and introduce E2Former-LSR, an equivariant transformer that explicitly integrates long-range attention blocks. E2Former-LSR exhibits stable error scaling, achieves superior fidelity in capturing non-covalent decay, and maintains precision on complex protein conformations. Crucially, its efficient design provides up to 30% speedup compared to purely local models. This work validates the necessity of non-local architectures for generalizable MLFFs, enabling high-fidelity molecular dynamics for large-scale chemical and biological systems.

Problem

• machine learning force fields
• long-range interactions
• macromolecular systems
• scalability
• non-covalent interactions

Innovation

• long-range interactions
• machine learning force fields
• equivariant transformer
• non-covalent interactions
• scalable molecular simulation