Research And Implementation Of Gene Pathway Enrichment Algorithm Based On Global And Local Information

In the field of biological research,high-throughput sequencing technology has spawned massive biological data.Starting from these biological data,using appropriate bioinformatics methods to explore biological pathways related to the pathogenesis of diseases and study the relationship between diseases and biological pathways is of great significance to the development of disease diagnosis and treatment technology.Gaining in-depth understanding of diseases from high-throughput massive data and discovering the complex mechanisms of diseases remains a challenge for researchers.Although in the past ten years,related researchers have developed some gene enrichment analysis methods to discover biological pathways related to diseases.However,these methods seldom consider the biological pathway information,or only consider the local information of the pathway,or the global information of the larger biological network,which do not conform to the actual existence of the pathway,which may lead to inaccurate results.Therefore,developing an effective gene pathway enrichment algorithm to discover disease-related biological pathways has become an important topic in computational biology.Genes in pathways affect biological processes such as diseases through complex interactions.Using the gene interaction data of the STRING database,this paper proposes a new method based on gene interactions that comprehensively considers the local information of genes in the pathway and the global information of the network.The pathway enrichment algorithm.The research in this paper mainly includes the following contents:(1)A pathway gene weighting model that fuses global and local information is constructed.This paper uses gene interaction data to generate a global biological network.Based on the improved TF-IDF algorithm,the local importance of the gene in the pathway and the global specificity in the pathway database were fused to define the influence of the gene,and a pathway gene weighting model was constructed.(2)A gene pathway enrichment algorithm that fuses global and local information of genes is proposed.This method fuses the gene weighting model with the traditional gene enrichment algorithm.Gene influence and phenotypic correlation values were fused together,then the enrichment score for the pathway was calculated,and finally statistically significant pathways were enriched by permuting genes.(3)A pathway enrichment analysis website was built.In order to facilitate users to use this algorithm for gene pathway enrichment analysis,this paper uses the algorithm to develop a gene pathway enrichment analysis website based on the SSM framework.The website includes user information module,pathway enrichment algorithm module and extended function module.The user information module is for the convenience of users to view their own enrichment results.The pathway enrichment algorithm module includes two enrichment modes: native pathway enrichment mode and pre-sorted pathway enrichment mode.Extended functional modules include functions such as viewing pathway gene interaction networks.In summary,this paper constructs a gene weighted model by extracting gene interaction data,and applies the model to gene pathway enrichment analysis,and finally builds a pathway enrichment analysis website.In this paper,the algorithm is applied to the enrichment of related pathways in hepatocellular carcinoma,colorectal cancer and lung cancer.Compared with the gene set enrichment analysis(GSEA)method,the method of calculating gene influence by fusing global and local information can enrich new related pathways,and exclude some irrelevant pathways,thereby improving the enrichment effect of disease-related pathways.