Accurate quantum chemical property prediction with transferable electron-density representations
Abstract Machine-learning models of molecules usually encode nuclear geometry, while spatial electronic information is either omitted or requires costly molecule-specific quantum calculations. Here we present eMol, an electron-density-driven model that overcomes this trade-off by coupling atomic scaffolds with density-derived probes via a retrievable memory, allowing compatible electronic environments to be reused across molecules without repeated first-principles density calculations. To integrate the transferred electronic information with nuclear geometry, eMol employs an equivariant atom–electron architecture that captures atomic geometry, local electronic structure and higher-order many-body interactions within a unified representation. To avoid computing electron density for every query, eMol uses structure-only retrieval and geometric alignment to reuse probes from a memory of approximately two million previously characterized molecules. This design yields consistently strong performance across 24 quantum-chemical property prediction tasks and 24 molecular potential-learning systems, and reproduces reference structural distributions in biomolecular dynamics. Its learned attention patterns align with local electron density, and it prioritizes literature-supported compounds in drug-activity datasets. More broadly, decoupling expensive physical information from real-time computation provides a generalizable paradigm for molecular learning.
Authors
- Hongxin Xiang (ORCID: https://orcid.org/0000-0001-8345-8735)
- Xiangxiang Zeng (ORCID: https://orcid.org/0000-0003-1081-7658)
- Xuan Lin (ORCID: https://orcid.org/0000-0001-7216-1821)
- Jun Xia (ORCID: https://orcid.org/0000-0002-7993-0803)
- Tong Wang (ORCID: https://orcid.org/0000-0002-9483-0050)
- Yang Li (ORCID: https://orcid.org/0000-0002-1848-212X)
- Yang Sun (ORCID: https://orcid.org/0009-0004-9501-3599)
- Jingyu Sheng
- Tianhang Zhang
- Dapeng Xiong (ORCID: https://orcid.org/0000-0002-9337-3839)
- Shuting Jin
- Qingrui Liu
Institutions
- Hunan University (CN)
- Wuhan University of Science and Technology (CN)
- The Hong Kong University of Science and Technology (Guangzhou) (CN)
- Xiangtan University (CN)
- Southeast University (CN)
- Tsinghua University (CN)
Publication Details
- Journal
- Research Square
- Published
- 2026-11-12
- DOI
- https://doi.org/10.21203/rs.3.rs-11439055/v1
- Primary Topic
- Machine Learning in Materials Science
- Type
- preprint