Research And Application Of Gene Function Prediction Based On Random Walk

In recent years,the study of gene has attracted much attention.Clinical medicine and biological experiments produce a large amount of biological data.At present,there are many databases that record the ontology data of genes and diseases,but most of them are very specific and cannot effectively discover the potential information of genes according to the association relationship of different genes.This topic fuses multiple gene-related data sources,uses the improved random walk algorithm to research and develop the gene integrated resource search system,and uses the integrated resources of multi-data source fusion in the system to carry out gene data mining,and discovers the potential function of genes.The experimental results show that the method proposed in this thesis can still achieve better prediction results after the fusion of multiple data sources.The main work of this thesis is as follows:1.This thesis combines several biological database resources,such as gene ontology,disease ontology and human phenotype ontology,constructs a gene-term association network by calculating the similarity between genes and terms,and visualizes the constructed association network;2.A weighted biased random walk algorithm based on multi-data source fusion is proposed to predict gene function.Based on the gene-term association network,the biased random walk algorithm was used to mine the structure information of the network to predict the gene function.3.A gene integrated resource search system based on the fusion of multiple data sources is developed.The system integrates the relevant data of multiple databases,provides a unified entrance for the access of multiple databases,and speeds up the speed of different gene retrieval.At the same time,the search results can be displayed in different views.