Integrative Analysis of Transcriptome and Methylation Data in Human Non-Small Cell Lung Cancer
Human lung cancer is the most prevalent cancer worldwide that consisting of two main subtypes: the non-small cell lung cancer (NSCLC) and the small cell lung cancer (SCLC). NSCLC comprises over 80% of lung cancer and the treatment of NSCLC is mostly guided by tumor stage, although distinctive molecular characteristics between two major subtypes of NSCLC, i.e., lung adenocarcinoma (LUAD) and squamous cell lung carcinoma (LUSC), have been increasingly identified. In this study, we integrated the gene expression data and methylation data to investigate the genetic differences between LUAD and LUSC. We further applied the Boruta package to select key features from LUAD and LUSC tumor samples to build predictive models of tumor stage. We finally obtained 6 key gene expression features and 4 key methylation features that can be reliably used in prediction of LUAD and LUSC stage.