Active learning for nonadiabatic molecular dynamics

Mikołaj Martyka, Lina Zhang, Fuchun Ge, Yi-Fan Hou, Joanna Jankowska*, Mario Barbatti*, Pavlo O. Dral*. Charting electronic-state manifolds across molecules with multi-state learning and gap-driven dynamics via efficient and robust active learning. npj Comput. Mater. 2025, 11, 132. DOI: 10.1038/s41524-025-01636-z. Preprint on ChemRxiv: https://doi.org/10.26434/chemrxiv-2024-dtc1w (2024-08-06).

Code and data

• Active learning with multi-state ANI, multi-state ANI model, and gapMD are implemented in the open-source MLatom (version >=3.10). Please see the MLatom page for more details. Links to tutorials are given below.
• All data is available in the above repository, with the exception of the azobenzene training set, which is available at the following DOI: 10.6084/m9.figshare.27024196
• ML models are provided as compressed .npz files, loadable with MLatom version >=3.10.
• Training sets are provided as xz compressed .json molecular databases, compatible with MLatom.

Tutorials

• active learning: https://xacs.xmu.edu.cn/docs/mlatom/tutorial_namd.html#active-learning
• multi-state ANI: https://xacs.xmu.edu.cn/docs/mlatom/tutorial_namd.html#multi-state-ani-models
• gapMD: https://xacs.xmu.edu.cn/docs/mlatom/tutorial_namd.html#gapmd