Using the SVM Method for Lung Adenocarcinoma Prognosis Based on Expression Level

Lung cancer is the deadliest cancer in the word, leading to over a quarter of death in the United States in 2017. Gaining precise information on cancer prognosis for patients would greatly benefit their decision making for further treatment plans. While previous studies tend to use histology information and genomic signatures for cancer prognosis, this study explores the possibility of using expression level alone to predict prognosis. Using over 200 patients from publicly available datasets with overall survival length and transcriptomic information, we use support vector machines to predict prognosis. Our result proves the effectiveness of such methodology, encouraging transcriptomic data to be collected for patients routinely if possible given the decreasing cost of RNA-Seq.