Novel molecular similarity search algorithm based on graph structure attention

Abstract

The invention provides a novel molecular similarity search algorithm based on graph structure attention, which includes the following steps: 1) with similar molecules as a pretraining data, expressingthe similar molecules into the format of molecular graph through a graph convolution neural network, and converting the molecular graph into feature vectors by introducing the attention mechanism into the graph convolution neural network; 2) training the network to obtain the parameters in the formula, thus generating the similarity calculation formula; 3) expressing the new molecular by the formula of the new molecular graph through the method in the step 1), and converting the new molecular graph into a new feature vector, and performing similarity calculation by using the similarity calculation formula to obtain similarity score. By means of the method, the accuracy of the molecular similarity calculation is improved on the basis of guaranteeing the universality of searching.

Description

technical field [0001] The invention belongs to the field of molecular similarity search, in particular to a novel molecular similarity search algorithm based on graph structure attention. Background technique [0002] At present, with the rapid development of information technology, computers play an increasingly important role in information technology in the field of drug discovery. In recent years, the development of artificial intelligence has brought more possibilities for drug discovery. Molecular similarity search is an important technology in drug small molecule discovery engineering. It has laid a theoretical foundation for many existing virtual screening technologies. Molecules are a commonly used screening tool. [0003] In chemical analysis, the analysis of physical and chemical properties based on similarity, the principle of similar miscibility, etc. are all based on [0004] "Compounds with similar structures generally have similar physical and chemical pr...

AI Technical Summary

Problems solved by technology

However, due to the huge amount of data in the molecular database, a molecule needs to be subjected to exponentially-sized incremental experiments, which greatly increases the calculation time and limits the speed of using the shape recognition method.

Images