SchNarc Databases for Molecular Motor Photoisomerization

# SchNarc Databases for Molecular Motor Photoisomerization This repository provides the SchNarc `.db` datasets used in our study of machine-learning-accelerated nonadiabatic dynamics and vibrational control in an overcrowded-alkene molecular motor. The repository contains two database files: - `PES.db`: dataset for training multi-state machine-learned potential energy surfaces used in excited-state and nonadiabatic dynamics simulations.- `Dipole.db`: dataset for training machine-learned dipole models used to describe light–matter interactions and field-driven dynamics. The datasets were constructed from molecular configurations sampled by static scans, Wigner sampling, short-time surface-hopping trajectories, adaptive sampling, and umbrella sampling, with explicit coverage of stable isomers, reactive regions, and conical-intersection-related geometries. If you use these databases, please cite the associated manuscript.