Identification The Gene And Mechanism In The Lung Cancer

Lung cancer is currently the most common cause of cancer death. About 130 million people die of lung cancer every year in the word.5-year survival rate of patients with lung cancer has remained unchanged at 14% for many years. In the past 20 years, high-throughput technology for lung cancer produced a large amount of data. It potentially provides the resources for lung cancer molecular mechanisms. Human lung cancer database HLungDB is a lung cancer platform related to genes, proteins and miRNA and corresponding clinical data. The main function of the platform is to provide a comprehensive cancer related network resources to promote the progress of lung cancer mechanism research. We manually extracted the molecular and human lung cancer data from the literature. Currently, we collected 2585 genes and 212 miRNA related data at different stages of lung cancer by text mining. In addition, we also integrated the transcription factor, transcription factor binding sites, promoter, SNP and other information. Since epigenetic regulation plays an important role in lung cancer, we also collected relevant data. We hope this database could enrich our biological knowledge about lung cancer and ultimately lead to new treatment for lung cancer. HLungDB database is available through (http://www.megabionet.org/bio/hlung). Next, we explore the potential mechanisms related to lung cancer from the point of view in the biological pathway cross-talk by integrating the microarray data and HLungDB data. By using factor analysis and hyper-geometric distribution methods, we analyzed the expression profile data and the biological pathway data to obtain significant lung cancer pathways and quantify the interaction degree between the pathways, then constructed a network of the lung cancer pathways. We also analyzed the overrepresentation of the pathways and detected the potential key genes involving cross-talk between pathways. As a result, we identified lung cancer related pathways which have been widely studied, such as ECM-receptor interaction pathway, p53 signaling pathway, cell adhesion molecules (CAMs) pathway, focal adhesion pathway, cell cycle pathway. In addition, we also detected a new lung cancer related pathway, complement and coagulation cascades. Many indirect evidences indicate that the new identified pathway indeed closely link to lung cancer. The lung cancer biological pathway network also demonstrated some sub-networks associated with lung cancer. By studying the overrepresentated pathways in the network, we proposed a hypothesis that cross-talk takes place among cell cycle, p53 signaling pathway and MAPK signaling pathway through gene GADD45b in coordination with other proteins. Our results related to lung cancer tumorigenesis can no doubt facilitate future lung cancer investigations.