The development of molecularly targeted drugs, specifically those which modulate the activities of one or several proteins involved in the pathogenesis of a cancer, is the most exciting field for cancer treatment because targeted anticancer drugs have the potential to provide dramatic clinical benefits with little toxicity. In order to develop new molecularly targeted drugs for lung cancer, the leading cause of cancer in the world, we have collected a large amount of data, including genetic/epigenetic (mutations, copy number variation, and methylation), mRNA expression, protein expression and genome-wide RNAi functional screening data on 108 non-small cell lung cancer (NSCLC) cell lines. Integrating these large-scale and complementary datasets from different sources will provide great opportunities to discover new molecular mechanisms of lung cancer. In Aim 1 of this study, we will develop a powerful computational model to integrate multiple genomic, proteomic and functional datasets to identify new lung cancer driver genes. Only a small subset of tumor driver genes is traditionally "druggable" targets. In Aim 2 of this study, we will use a data-drive and unbiased approach to discover and evaluate potential new therapeutic targets in lung cancer. A novel reverse engineering approach will be proposed to construct a lung-cancer-specific gene network. In Aim 3 of this study, we will develop a publicly available comprehensive lung cancer database with a user-friendly interface and powerful analysis engine. This database will include all genomic, proteomic and functional data together with the de-identified clinical data used in this study. By using the state-of-the-art information technolog, we will integrate these datasets with analytic algorithms and a user-friendly interface in a publicly available database so that researchers worldwide can utilize and test the data and computational tools generated from this study.