Quantum chemistry calculation and machine learning prediction ofelectric dipole in complex system

摘要

Based on density functional theory calculations,we proposed that electric dipole,aparameter deduced from the distribution of ground or excited electronic states,is a usefuldescriptor to describe and evaluate the structure-property relationship of complex systems.An efficient algorithm was developed to simulate the photo-response of bio-complexes withtheir electric dipoles extracted from combined quantum mechanics and molecularmechanics (QM/MM) calculations.