Learning Non-Local Molecular Interactions via Equivariant Local Representations and Charge Equilibration

2025-01-31

AI Summary

Conventional graph neural networks (GNNs) rely on local potential functions constrained by chemical locality, limiting their ability to model long-range, non-local intermolecular interactions—such as charge transfer, electrostatics, and dispersion effects. Method: This work introduces the CELLI (Charge Equilibration Layer), the first physics-informed integration of the Qeq charge equilibration method into an equivariant GNN framework. CELLI combines equivariant local representations with Qeq theory and fourth-generation high-dimensional neural networks (4GHDNN) to construct an efficient, differentiable long-range interaction module that is model-agnostic and scalable. Results: Evaluated across multiple benchmark datasets and large molecular systems, CELLI achieves accuracy comparable to message-passing neural networks (MPNNs) while delivering ~2× computational speedup. It significantly enhances the Allegro architecture’s capacity to capture dynamic charge redistribution and other non-local electronic phenomena, establishing a new paradigm for scalable, physically grounded molecular modeling.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches 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

Graph Neural Network (GNN) potentials relying on chemical locality offer near-quantum mechanical accuracy at significantly reduced computational costs. By propagating local information to distance particles, Message-passing neural networks (MPNNs) extend the locality concept to model interactions beyond their local neighborhood. Still, this locality precludes modeling long-range effects, such as charge transfer, electrostatic interactions, and dispersion effects, which are critical to adequately describe many real-world systems. In this work, we propose the Charge Equilibration Layer for Long-range Interactions (CELLI) to address the challenging modeling of non-local interactions and the high computational cost of MPNNs. This novel architecture generalizes the fourth-generation high-dimensional neural network (4GHDNN) concept, integrating the charge equilibration (Qeq) method into a model-agnostic building block for modern equivariant GNN potentials. A series of benchmarks show that CELLI can extend the strictly local Allegro architecture to model highly non-local interactions and charge transfer. Our architecture generalizes to diverse datasets and large structures, achieving an accuracy comparable to MPNNs at about twice the computational efficiency.

Problem

• Graph Neural Networks
• Long-range Effects
• Molecular Interactions

Innovation

• Charge-Equilibrium Long-range Layers (CELLI)
• Graph Neural Networks (GNNs)
• Efficient Long-range Interactions