Common subcluster mining to explore molecular markers of lung cancer

Abstract

We present a methodology, common subcluster mining, to explore gene expression data for possible biomarkers of lung cancer. Subclusters refer to the peaks formed through superimposition of clusters obtained from expression data of normal samples. Application of the method on the corresponding data sets from diseased samples extracts the genes that undergo high fold changes. The potential candidate genes are examined on the datasets of Stage I through stage IV of the disease. Few genes emerge as indicative molecular markers of lung cancer.