Breast Cancer Research Based On Differential Network

Malignant tumors penetrate healthy body tissues and destroy them,seriously threatening human life.Breast cancer,a malignancy that develops from breast cells,is the most common cancer among women in almost all regions of the world,accounting for a third of cancer diagnoses.For complex diseases such as breast cancer,it is necessary to make great efforts to obtain data on different conditions and to construct static networks to explain different gene regulation mechanisms.However,since the regulatory network of genes is dynamic and specific to the environment,it is the most important work to analyze the topological changes of the network and to construct the analysis model of breast cancer genes based on the differential network.The data studied in this paper are from GEO and TCGA databases,among which the data in TCGA is the main data of the study.Firstly,gene differential expression analysis and PPI analysis were performed on the data in GEO,among which GO＿BP and KEGG pathway analysis were used to find out the biological processes and pathways in which differential genes are highly enriched,and PPI analysis was used to obtain proteins with strong interactions.Secondly,survival and mutation analysis were performed on mi RNA,clinical data in TCGA and mutation data of BRCA in TCGA-BRCA project.In the survival analysis,the edge R method was used to search for differential genes,the KM method was used to estimate the survival rate at each time and draw the survival curve,and the Cox regression model was used to study and model the factors affecting survival(including its size and direction).Mutation analysis of breast cancer genes using maftools,in which Fisher’s exact test was used to analyze mutated genes’ mutual exclusion and co-occurrence,and Co MEt test was used to find gene sets containing more than two mutated genes.Finally,653 samples including breast cancer tissues and adjacent normal samples were studied by DINGO and i DINGO analysis models.Among them,GLasso obtained sparse estimation of precision matrix by maximizing the penalty log likelihood ratio,Greg.em function fitted precision regression model,Sigma function obtained group specific covariance matrix from the parameters of the two analysis models,the bootstrap algorithm for genetic differences in scores,plot Network function drawn DINGO and i DINGO network identification of differences.In this paper,PPI analysis shows that the interaction between AKR1C3 and AKR1C1 is the strongest,survival analysis shows that age has the most significant impact on the survival rate of breast cancer patients,mutation analysis shows that gene TP53,GATA3,CDH1 are mutually exclusive,DINGO analysis shows that GSK3 protein was identified as hubs in different network,DINGO and i DINGO show that the edge value of MIR214 gene is the highest.These genes and proteins are involved in the development of breast cancer,so the above analysis methods are conducive to the study of breast cancer.