A method for predicting disease-related metabolites by adopting double random walk

Abstract

The invention discloses a method for predicting disease-related metabolites by adopting double random walk. The method comprises the following steps: firstly, calculating semantic similarity of a disease by utilizing the semantic information of the disease, and calculating functional similarity of the metabolites according to the relationship between the disease and the metabolites; secondly, according to the known metabolite-disease relationship, calculating Gaussian kernel action spectrum similarity of the disease and the metabolites respectively; thirdly, constructing a final disease similarity network based on the semantic similarity of the disease and the Gaussian kernel action spectrum similarity of the disease, and constructing a final metabolite similarity network based on the function similarity of the metabolites and the Gaussian kernel action spectrum similarity of the metabolites; and finally, constructing a two-part heterogeneous network based on the disease similarity, the metabolite similarity and the known metabolite-disease relationship network, and predicting the disease-related metabolites in the heterogeneous network by utilizing a double random walk method. Tests prove that the method disclosed by the invention can predict the disease-related metabolites, provides a basis for biological experiments and improves diagnosis and treatment level of the disease.

Description

technical field [0001] The invention belongs to the field of biological information, and in particular relates to a method for predicting disease-related metabolites by using double random walks. Background technique [0002] Metabolites are the end products of cellular regulatory processes and are often considered the final responses of biological systems to genetic or environmental changes. Metabolites play an important role in the maintenance, growth and reproduction of organisms, and the level of metabolites can directly reflect the physiological state of the human body. There has been sufficient evidence that diseases are always accompanied by changes in metabolites. Therefore, identifying abnormalities and identifying disease-associated metabolites is of great significance not only for improving clinical diagnosis, but also for better understanding metabolic pathological processes. [0003] With the development of molecular technology, more and more researchers are at...

AI Technical Summary

Problems solved by technology

RWR is based on the idea that metabolites with high similarity are very likely to be related to the same disease. It mainly walks on the metabolic network and cannot fully combine the information of metabolic network, disease network and disease metabolite association network. Cannot predict disease-associated metabolites with no known relationship

Although some achievements have been made, the results are not enough to guide biological experiments

Images