Forces are not Enough: Benchmark and Critical Evaluation for Machine Learning Force Fields with Molecular Simulations

arXiv (Cornell University) · 2022 · 162 citations

• Computer Science
• Computer Science
• Machine Learning

Molecular dynamics (MD) simulation techniques are widely used for various natural science applications. Increasingly, machine learning (ML) force field (FF) models begin to replace ab-initio simulations by predicting forces directly from atomic structures. Despite significant progress in this area, such techniques are primarily benchmarked by their force/energy prediction errors, even though the practical use case would be to produce realistic MD trajectories. We aim to fill this gap by introducing a novel benchmark suite for learned MD simulation. We curate representative MD systems, including water, organic molecules, a peptide, and materials, and design evaluation metrics corresponding to the scientific objectives of respective systems. We benchmark a collection of state-of-the-art (SOTA) ML FF models and illustrate, in particular, how the commonly benchmarked force accuracy is not well aligned with relevant simulation metrics. We demonstrate when and how selected SOTA methods fail, along with offering directions for further improvement. Specifically, we identify stability as a key metric for ML models to improve. Our benchmark suite comes with a comprehensive open-source codebase for training and simulation with ML FFs to facilitate future work.

Hierarchically structured bioinspired nanocomposites

Nature Materials · 2022 · 537 citations

• Computer Science
• Nanotechnology
• Computer Science

Mesoscopic and multiscale modelling in materials

Nature Materials · 2021 · 370 citations

• Computer Science
• Computer Science
• Algorithm

Molecular insights into charged nanofiltration membranes: Structure, water transport, and water diffusion

Journal of Membrane Science · 2021 · 50 citations

• Chemistry
• Chemical engineering
• Chemical physics