Molecular property prediction method and system

Abstract

The invention provides a molecular property prediction method and system, and relates to the fields of quantum chemistry/computational chemistry, chemical informatics and machine learning/artificial intelligence. Under the framework of chemical multi-world explanation, a density functional theory, chemical informatics and machine learning/artificial intelligence means are used, information such asa molecular structure, a base group, a functional and the like is used as input, and a prediction result of molecular property is output through a machine learning model. According to the method, anytype of molecular structure and any calculation strategy can be predicted, and the method is more accurate than a common experience method and a regression analysis method.

Description

technical field [0001] The present invention relates to the fields of quantum chemistry / computational chemistry, cheminformatics, and machine learning / artificial intelligence, specifically a method using density functional theory, cheminformatics, and machine learning / artificial intelligence under the framework of the chemical multi-world theory. Theoretical methods for predicting the properties of unknown molecules. Background technique [0002] Calculation of various intrinsic properties of molecules is one of the core issues in quantum chemistry / computational chemistry. Early solutions were based on empirical and semi-empirical model Hamiltonian theory, such as Hückel molecular orbital theory and classical valence bond theory. Since the 1990s, with the rapid development of computer software and hardware, theoretical methods based on ab initio Hamiltonian have occupied a major position. These theoretical methods include Hartree-Fock self-consistent field theory and electr...

AI Technical Summary

Problems solved by technology

Therefore, the combination of basis sets and functionals in density functional calculations can be considered to be infinitely diverse, which also leads researchers to choose different basis sets and functionals for calculation and testing in the calculation process of related properties; at the same time, in a specific functional The calculation results obtained under the functional and basis set cannot be directly extrapolated to other combinations of functional and basis set

The above inconveniences have also greatly affected the work efficiency of scientific researchers in this area.

Images