This study developed a method to identify disease-correlated pathways by integrating copy numbers (CN) and gene expression (GE). To evaluate the correlation between CN and GE, a suitable window size was assessed by simulation. Gene Set Enrichment Analysis (GSEA) was utilized to identify the possible pathways by CN, GE, and their correlations, respectively. Each of those enriched pathways was further assigned a score to incorporate the information from CN, GE, and their correlations. A dataset of 44 female nonsmoking lung cancer patients with both normal and tumor tissues was used to evaluate the performance of this method. To further appraise the predicting abilities of those pathways, patients were classified by support vector machines using the pathways identified by only copy number, only gene expression and incorporating CN, GE, and their correlations. The results showed that the proposed method earned higher accuracy, sensitivity and specificity than traditional methods.
展开▼
