Multi-task drug screening method and system based on knowledge graph assistance

Abstract

The invention discloses a multi-task drug screening method based on a knowledge graph, and the method comprises the steps: obtaining an entity mapping table after all drugs and all compounds are mapped to a knowledge graph data set, a candidate drug set of all drugs, and a pre-trained multi-task prediction model, obtaining a question inputted by a user, and obtaining a candidate drug set of all drugs; predicting a target point x and a candidate drug set in the problem by using the obtained multi-task prediction model to obtain a drug probability score table and a compound probability score table, and calculating and merging the obtained drug probability score table and the compound probability score table by using a result integration algorithm to obtain a drug probability score table and a compound probability score table; and sorting the probability scores in the combined probability score table from large to small, and selecting entities corresponding to the first 20 drugs from a sorting result as a screening result and returning the screening result to the user. The problem that an existing prediction model based on a knowledge graph cannot consider the characteristics of drug molecules and proteins can be solved.

Description

technical field [0001] The invention belongs to the technical field of artificial intelligence, and more specifically relates to a multi-task drug screening method and system assisted by knowledge graphs. Background technique [0002] In the field of drug discovery, discovering new drugs is often extremely time-consuming and costly. In recent years, with the development of computer computing power and the completeness of drug-related data, the drug screening process using computers can greatly accelerate the process of new drug development and effectively reduce the cost of new drugs. In drug screening, discovering the interaction between the drug and the target or the interaction between the compound and the protein is a necessary part and the basis of drug screening. Therefore, existing methods will perform drug screening by predicting the interaction between the drug and the target. [0003] Existing drug screening methods are mainly machine learning-based methods and n...

AI Technical Summary

Problems solved by technology

[0005] In view of the above defects or improvement needs of the prior art, the present invention provides a multi-task drug screening method and system based on knowledge map assistance, the purpose of which is to solve the problem that the existing drug target prediction model cannot provide high predictive ability , and for the data sparse scene and the situation of false positive samples, the adaptability deviation leads to the problem of low prediction accuracy

Images