Method and device for extracting molecular fingerprints and calculating relevancy based on molecular fingerprints

Abstract

The invention discloses a molecular fingerprint extraction method and device and a relevancy calculation method and device based on the molecular fingerprint extraction method, and the molecular fingerprint extraction method comprises the steps: obtaining a plurality of characters of a to-be-detected molecule; according to the plurality of characters and a preset character dictionary, respectivelydetermining a feature vector corresponding to each character; and extracting the molecular fingerprint of the to-be-detected molecule according to the feature vector and the molecular fingerprint extraction model. By implementing the method and the device, the problem that the potential activity aspect of the molecule is not related even if the structure is relatively similar due to the fact thatthe molecular fingerprint determined on the basis of artificially designed molecular characteristics cannot describe the overall structure of the molecule in the related technology is solved, and thekey characteristic information of the molecule is obtained, and the relatively accurate molecular activity relevancy information is obtained, so the molecular similarity can be accurately evaluated,so ligand-based virtual screening can be more accurate and efficient, and the time required by virtual screening is effectively shortened.

Description

technical field [0001] The invention relates to the field of data processing and analysis, in particular to a method and device for extracting molecular fingerprints and calculating correlations based thereon. Background technique [0002] Finding active potential molecules is a critical part in the process of drug design and discovery, and active potential molecules can be named as HIT molecules. Usually, pharmaceutical and chemical experts will use computer and other related technologies to assist in speeding up the process of finding HIT molecules, and virtual screening is one of the most important technologies. Molecular fingerprints are generally used to determine the similarity between reference ligands and candidate ligands, that is, the virtual screening process of molecules. Molecular fingerprint is an abstract representation of molecules, which converts molecules into a string of bits, and compares molecules according to various calculation methods of vector simil...

AI Technical Summary

Problems solved by technology

[0005] Therefore, the technical problem to be solved by the present invention is to overcome the incomplete description of the overall structure of the molecule in the prior art, resulting in the selected molecules, even if they are similar in structure, but not close in terms of molecular potential activity. A method and device for extracting molecular fingerprints and calculating correlations based thereon

Images