Evolution of Network Biomarkers from Early to Late Stage Bladder Cancer Samples

Figure 1

The flowchart of constructing both stages of network marker of bladder cancer and the investigation of the carcinogenesis mechanisms. We integrate microarray data, GO database, and PPI information to construct the PPI network. These data are used for pool selection, and then the selected proteins and the microarray data are used for the contribution of protein-protein interaction network (PPIN) by maximum likelihood estimation and model order detection method, resulting in bladder cancer PPIN (CPPIN) and noncancer PPIN (NPPIN) of early and late stage. The two constructed PPINs can be used for the determination of significant proteins of tumorigenesis by the difference between two PPI matrices of two constructed PPINs. With the help of the differential PPI matrix (network) between CPPIN and NPPIN, carcinogenesis relevance value (CRV) is computed for each protein, and significant proteins in carcinogenesis are determined based on P value the CRVs of these proteins in the differential PPI matrix between CPPIN and NPPIN. These significant proteins are obtained for early and late stage bladder cancers.