EDBench: Large-Scale Electron Density Data for Molecular Modeling

2025-05-14

AI Summary

Existing molecular machine learning force fields (MLFFs) commonly neglect electron density (ED) modeling, despite ED being a fundamental variable in the Hohenberg–Kohn theorem that uniquely determines ground-state properties. High-fidelity ED data—essential for integrating ED into MLFFs—remain scarce due to the prohibitive computational cost of density functional theory (DFT) calculations. Method: We introduce EDBench, the first large-scale, high-accuracy ED benchmark dataset comprising 3.3 million molecules. We extend PCQM4Mv2 and augment it with DFT-computed ED labels, then develop an ED-driven multitask evaluation framework encompassing ED prediction, retrieval, and generation. A deep neural network architecture enables efficient, scalable ED modeling. Contribution/Results: Our model achieves ED prediction errors below 0.01 a.u., with inference speeds four orders of magnitude faster than DFT. This significantly enhances MLFFs’ understanding of electronic structure and improves generalization across diverse molecular systems.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion
• Humans and AI: Human-in-the-loop Machine Learning

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Existing molecular machine learning force fields (MLFFs) generally focus on the learning of atoms, molecules, and simple quantum chemical properties (such as energy and force), but ignore the importance of electron density (ED) $ ho(r)$ in accurately understanding molecular force fields (MFFs). ED describes the probability of finding electrons at specific locations around atoms or molecules, which uniquely determines all ground state properties (such as energy, molecular structure, etc.) of interactive multi-particle systems according to the Hohenberg-Kohn theorem. However, the calculation of ED relies on the time-consuming first-principles density functional theory (DFT) which leads to the lack of large-scale ED data and limits its application in MLFFs. In this paper, we introduce EDBench, a large-scale, high-quality dataset of ED designed to advance learning-based research at the electronic scale. Built upon the PCQM4Mv2, EDBench provides accurate ED data, covering 3.3 million molecules. To comprehensively evaluate the ability of models to understand and utilize electronic information, we design a suite of ED-centric benchmark tasks spanning prediction, retrieval, and generation. Our evaluation on several state-of-the-art methods demonstrates that learning from EDBench is not only feasible but also achieves high accuracy. Moreover, we show that learning-based method can efficiently calculate ED with comparable precision while significantly reducing the computational cost relative to traditional DFT calculations. All data and benchmarks from EDBench will be freely available, laying a robust foundation for ED-driven drug discovery and materials science.

Problem

• Lack of large-scale electron density data for MLFFs
• Time-consuming DFT calculations limit ED applications
• Need for ED-driven benchmarks in molecular modeling

Innovation

• Large-scale electron density dataset EDBench
• Benchmark tasks for ED prediction and generation
• Learning-based ED calculation reduces DFT cost