This lung cancer project has been instrumental in determining genetic markers that could inform clinicians of the potential success (or failure) of a particular treatment before chemotherapy begins. This type of information can be used to improve the quality of care and decrease costs as patients get the most beneficial treatment the first time or do not undergo a potentially ineffective treatment needlessly. Through this work, a 51 gene set that can predict chemotherapy resistance to two commonly used agents in patients with non-squamous cell lung cancer has been identified. Significant differences in genomic profiles were also found between squamous and non-squamous cell carcinomas that would encourage the division of the two subtypes in future studies to better reflect more accurate response/outcome to therapy. In addition predictive DNA signatures were found to be different than the prognostic signatures providing the basis for a 2500 patient, multi-center clinical trial that seeks to validate biomarkers to improve the early detection of lung cancer.