2026-10-11T15:21:04Z https://api.figshare.com/v2/oai

oai:figshare.com:article/33978177 2026-09-24T00:07:21Z category_1 category_873 category_734 category_931 portal_63 item_type_3 month_year_09_2026

MBAvol: An Interpretable Descriptor Scheme for Crystal-Density Modeling of HEDM Chao Chen (195669) Yiding Ma (1985614) Xiaoyan Wang (8433) Xiaokai He (19352274) Zhixiang Zhang (136083) Yingzhe Liu (1738708) Biophysics Chemical Sciences not elsewhere classified Biological Sciences not elsewhere classified Information Systems not elsewhere classified three design guidelines physically motivated atomic mean absolute error interpretable feature representation interpretable descriptor scheme hedm machine learning conventional models often eight candidate hedms guided hedm discovery density functional theory 3 </ sup cryst </ sub high predicted ρ hedms ), thereby reducing shap analysis sample accuracy predict properties practical framework performs well limited interpretability extrapolative capability energetic cocrystals efficient route density prediction density modeling density materials data required analytical fitting accurate modeling 026 g Machine learning (ML) offers an efficient route to predict properties of high-energy-density materials (HEDMs), but conventional models often have limited interpretability and out-of-sample accuracy. Here, we introduce the multibody approximation volume descriptor (MBAvol) for crystal-density (ρ_cryst) prediction. MBAvol represents molecular spatial information through physically motivated atomic and multibody volume terms derived from density functional theory and analytical fitting, thereby reducing the data required for accurate modeling. The resulting regression model achieves a mean absolute error of 0.026 g/cm³ and performs well for energetic cocrystals. SHAP analysis of the MBAvol terms provides quantitative, chemically interpretable structure–density relationships, from which three design guidelines were derived. Guided by these relationships, eight candidate HEDMs with high predicted ρ_cryst were identified. MBAvol therefore combines data efficiency, extrapolative capability, and interpretable feature representation, providing a practical framework for crystal-density prediction and structure-guided HEDM discovery. 2026-09-23T00:00:00Z Dataset Dataset 10.1021/acs.jpclett.6c02077.s001 https://figshare.com/articles/dataset/MBAvol_An_Interpretable_Descriptor_Scheme_for_Crystal-Density_Modeling_of_HEDM/33978177 CC BY-NC 4.0