A tool for analyzing difference of gene expression in a pathway

Research on pathway has been much appealing for studying the biological biomarkers, cell mechanism, and cancer proliferation in recent years. Pathway has been typically related to human diseases for drug discovery especially for tumor-targeted therapy. Since the scanty of the knowledge about cancer-related pathways, further exploration is required. Several machine-learning methodologies have been used to construct the prediction model of signaling pathway with protein-protein interactions (PPIs) and microarray gene expression data. However the gene expression profile of genes in a pathway remains underdeveloped. In this article, a pathway-based gene expression data analysis and visual tool is designed. The tool could be applied to illustrate the patterns of gene expression in different conditions. We obtained ten human cancer-related signaling pathways and three gene expression datasets (colorectal cancer, cervical cancer and breast cancer) from public resources on the Internet to present the utility of this pathway-based analysis tool. Furthermore, these expression profiles in control and abnormal status were compared to discuss the feasibility for inferring pathways from gene expressions. There are discrepancies of gene expressions in a pathway among different situations. Hence, researchers should contemplate more cautiously on applying gene expressions to predict path-ways.