Research On Identification Of Biomarkers For Breast Cancer Based On Coexpressed Network

Cancer is seriously endangering human health.Breast cancer is the second largest cancer in the world.Since the late 1970 s,the global incidence of breast cancer has been on the rise,seriously threatening the health of women.At present,although some achievements have been made in breast cancer treatment and drug research,there is not a very effective treatment method due to lack of understanding of the pathogenesis of such complex diseases.In recent years,with the development of immunology,molecular biology and genomics techniques,identification of valuable biomarkers has become a hot topic in current research.Over the past two decades,global gene expression profiles have become one of the common tools for the study of complex diseases.For example,Pokemon accelerates the development of hepatocellular carcinoma by regulating Akt and ERK-mediated cell signaling pathways.Propofol inhibits the metastasis of cancer cells by activating the PI3 K signaling pathway.Compared with the gene expression differential analysis,differential correlation and differential co-expression analysis have a more profound understanding.Therefore,we should not only consider the change of single gene expression value,we should find the key genes that regulate life activities from network.In this paper,we have a research on identification of biomarkers for breast cancer based on coexpressed network.The main achievements and innovations are summarized as follows: The existing biomarker identification methods have some shortcomings.For example,the gene expression data are also obtained by manual collation,when selecting the differential genes only consider the gene expression values,ignoring the regulatory between genes.When using WGCNA to construct co-expression network,there isn't a filtering process.Then the results will be affected.Based on these problems,this paper presents a new experimental framework for the identification of breast cancer biomarkers,which effectively combines the SAM algorithm with the WGCNA algorithm.Firstly,SAM was used to screen the differentially expressed genes,and then the correlation coefficient between gene pairs was calculated for the differentially expressed genes.By using the characteristics of the differentiated network,the genes that were greatly changed in the correlation between gene pairs under different experimental conditions were screened out,And then use these genes to construct a weighted co-expression network with WGCNA.The network is divided into modules by dynamic hierarchical clustering,and then biological analysis is performed on each module to find biologically meaningful breast cancer biomarkers,and finally we got a biomarker of 20 genes.The experimental framework proposed in this paper not only considers the characteristics of large amount of biological data and noise,but also solves the requirement of WGCNA algorithm for the amount of input data,not only reduces the load of computer memory,saves time,but also improves the accuracy of module and biomarker.As for the biomarker,it is verified from the biological significance and the classifier,the result is good.Using these 20 genes as characteristics,classifying quality of normal samples and diseased samples were higher than those of known biomarkers,which indicated that the biomarkers identified in this study were reliable,and the results showed that the algorithm proposed in this paper is a effective complex disease recognition algorithm.In order to verify the feasibility of theoretical research,this paper implements a breast cancer biomarker recognition system.When realizing the system,we use R to process the biological data and related algorithms based on a B/S architecture.And associate the front-end page display with backend data processing by using java web.There are many attributes of gene data to consider.Therefore,adding more and more comprehensive factors,such as GO data and pathway data,to construct a dynamic co-expression network in order to further explore the pathogenic markers associated with breast cancer will be the focus of future research.