An integrative method of mining genes related to complex disease like Clear Cell Renal Cell Carcinoma with mRNA microarray data

Abstract

The goal of genomics research is to describe the network of molecules and interactions between genes and disease processes in cells. Due to the structural of a network play a key role for deciphering the cellular function. Therefore, the analysis of the structure of gene networks for tissues with and without disease can provide better comprehending of the formation and development of disease on the molecular level. In this paper, the Wilcoxon rank-sum test method is carried out on gene expression profiles in kidney tissues with and without clear cell renal cell carcinoma (ccRCC) and to obtain candidate genes. Furthermore, mutual information networks of these genes are constructed. Base on the seven statistics of the two network, a new method to filter structural key genes is proposed that may have significant impacts on the development of diseases, and twenty two genes are selected as potential pathogenic key genes of ccRCC. Empirical studies on cancer show that they are cancer-related genes which respect the efficient of the method. At the same time, ten of these genes are closely related to the formation and development of ccRCC leaving the other genes open. Furthermore, we predict two pathways which may play an important role of development of ccRCC based on GO annotation. These predictions call for more empirical studies on the functions of the gene in clear cell renal cell carcinoma.