Discriminative microarray analysis reveals a global reprogramming of transcriptional profile in the human lung adenocarcinoma and predicts diagnostic and prognostic biomarkers

Lung cancer is the leading cause of cancer death. The most common type of lung cancer is lung adenocarcinoma (AC). The genetic mechanisms of the early stages and lung AC progression steps are poorly understood. There is currently no clinically applicable gene test for the early diagnosis and prediction of lung AC. However using massive microarray analysis of the primary tumors and adjacent tissues many essential diagnostic and prognostic biomarkers of AC could be discovered and clinically used. We propose a feature section method that allowed us to identify a several sub-sets of the most extreme discriminative variables (e.g. expressed genes), given separate samples (e.g. tissue types) in each selected pair without tissue type misclassification errors. We identified the global reprogramming of the transcriptome in human lung AC tissue versus normal lung tissue, which is associated with more than 2500 genes discriminating the tissues at 100% accuracy. Clustering analysis and GO annotation tools found that the up-regulated genes in lung AC are significantly associated with the mitotic cell cycle; inversely, strong silencing of lung tissue-specific, cell communication, motility and the immune system genes in AC vs normal lung tissue was found. A short list of prospective early AC diagnostic biomarkers/signatures including genes of the mitosis, cell adhesion, and drug resistance family is reported, validated and discussed in the context of a concept of stratified medicine.