Enhancing molecular dynamics with equivariant machine-learned densities

2026-04-27

AI Summary

Traditional interatomic potential models struggle to predict electronic observables such as dipole moments and polarizabilities. This work proposes DenSNet, a novel approach that uniquely targets the ground-state electron density as its central learning objective. By employing an SE(3)-equivariant neural network to model the Hohenberg–Kohn map and integrating atom-centered Gaussian basis expansions with a Δ-learning strategy, DenSNet enables unified prediction of both energy and electronic structure directly from atomic configurations. The method is validated on ethanol, ethanethiol, resorcinol, and oligothiophenes, yielding infrared spectra in excellent agreement with density functional theory (DFT) calculations or experimental data. Furthermore, it successfully extrapolates to long-timescale molecular dynamics simulations of a 12-mer, demonstrating transferable accuracy in predicting electronic properties and spectroscopic responses.

Technology Category

• Machine Learning: Kernel Methods
• Search and Optimization: Learning to Search
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Machine-learning interatomic potentials (MLIPs) have enabled molecular dynamics at near ab initio accuracy, yet remain limited to energies and forces by construction, leaving electronic observables such as dipole moments and polarizabilities inaccessible. We introduce DenSNet, a density-first approach to machine-learned electronic structure that learns the Hohenberg--Kohn map from nuclear configurations to the ground-state electron density. Our approach employs an SE(3)-equivariant neural network to predict density coefficients of a flexible atom-centered Gaussian basis, combined with a $Δ$-learning strategy that uses superposed atomic densities as a prior to accelerate training. A second equivariant network then maps the predicted density to the total energy, providing a unified framework for molecular dynamics and electronic structure. We validate DenSNet on ethanol, ethanethiol, and resorcinol, where infrared spectra from machine-learned trajectories show excellent agreement with experimental gas-phase measurements. To test scalability, we train on polythiophene oligomers with 1--6 monomers and extrapolate to chains of up to 12 monomers, generating stable long-time trajectories whose infrared spectra agree with reference density functional theory calculations. Here, we show that reinstating the electron density as the central learned quantity opens a practical route to transferable prediction of spectroscopic and electronic observables in large-scale molecular simulations.

Problem

• machine-learning interatomic potentials
• electron density
• electronic observables
• molecular dynamics
• spectroscopic properties

Innovation

• equivariant neural networks
• electron density
• machine-learned interatomic potentials
• Δ-learning
• molecular dynamics