Screening And Prognostic Evaluation Of Breast Cancer Biomarkers

In recent years,the incidence rate of breast cancer has increased rapidly,and has ranked first place in female malignant tumors.Breast cancer is a genetic disease.Mining effective information about biological data is an important way to study biomarkers of the disease.Many gene mutations and gene expression disorders are involved in the complex biomolecular networks of tumor patients.Therefore,the regulation mechanism and metabolism mechanism of biomolecules can be efficiently studied,through the effective extraction of tumor biological information.This project provides a new way to explore the occurrence and prognosis of breast cancer,through screening biomarkers and studying their biological functions in breast cancer.The main contents of this article are as follows:(1)Abnormal expression of proto-oncogenes or tumor suppressor genes is caused by gene mutation.Among them,single nucleotide variation(SNV)is closely related to the pathogenesis of tumor diseases.Firstly,the key mutation gene is obtained through social network algorithm,they are respectively PIK3 CA,TP53,CDH1 and GATA3.SNV law is analyzed by the electron-ion interaction pseudopotential(EIIP).At the same time,the relationship between key gene mutations and patient prognosis was validated,based on overall survival.The results show that the lower the base's EIIP value,the more prone to SNV;On the other hand,the prognosis of breast cancer patients can be better predicted by TP53,PIK3CA(CI = 95%,P<0.05).(2)Breast cancer as a genetic disease.The prognosis of patients can be affected by abnormal gene expression.The prognosis related genes involved in the biological process can be mined out,through the biological analysis of these genes.In this chapter,299 candidate genes were obtained by topological analysis of protein-protein interaction(PPI)network.Then,key genes were obtained by lasso algorithm and the risk model was established.The AUC value of the model was 0.905.Through bioaccumulation analysis,it can be found that the risk model related genes(RMRG)participate in the biological process of breast cancer cell proliferation,genetic material and protein synthesis.Therefore,these RMRG can be used as potential biomarkers of breast cancer prognosis.(3)Competitive endogenous RNA(ceRNA),as a biomarker and potential therapeutic target,has shown great research value and clinical application prospect in exploring the pathogenesis of tumor.In this chapter,the ceRNA network of breast cancer was analyzed systematically.First,the weights of the nodes in the ceRNA network are calculated by using the edge clustering coefficient(ECC)and pearson correlation coefficient(PCC).Then,LINC00466,CHL1-AS2 and LINC00337 were selected as biomarkers of breast cancer by the method of stepwise feature selection based on random forest(SFS-RF).The results showed that these RNAs showed good performance in the recognition of breast cancer samples.Among them,LINC00466 and CHL1-AS2 are closely related to the prognosis of breast cancer patients.