Integrated analyses of copy number variations and gene expression in lung adenocarcinoma.

摘要: Numerous efforts have been made to elucidate the etiology and improve treatment of lung cancer, but overall five-year survival rate is still only 15%. Identification prognostic biomarkers for cancer using gene expression microarrays poses a major challenge in that very few overlapping genes reported among different studies. To address this issue, we performed concurrent genome-wide analyses copy number variation identify reproducibly associated with tumorigenesis non-smoking female adenocarcinoma. The genomic landscape frequent variable regions (CNVRs) at least 30% samples was revealed, their aberration patterns were highly similar several studies previously. Further statistical analysis located CNVRs identified 475 differentially expressed between tumor normal tissues (p<10(-5)). We demonstrated reproducibility these another study (p = 0.0034, Fisher's exact test), showed concordance variations changes by elevated Pearson correlation coefficients. Pathway revealed two dysregulated functions tumorigenesis: regulation via AKT signaling cytoskeleton reorganization. validation enriched pathways three independent cohorts effective prediction survival. In conclusion, integrating profiles variations, genes/pathways may serve as tumorigenesis.