Identifying Cancer Subnetwork Markers Using Game Theory Method

In this paper, a novel game theory method is proposed to identify subnetwork markers by integrating gene expression profile and protein-protein interaction network. The proposed method has been evaluated on different cancer datasets in order to classify cancer phenotypes. To verify the performance of our approach, the identified subnetwork markers are compared with a greedy search method. The proposed method is capable of identifying robust subnetwork markers and presents higher classification performance.