A Network-based Approach to Classify Disease Stages of Prostate Cancer Using Quantitative Network Measures

Abstract-- This study deals with investigating the classification performance of information-theoretic measures when applied to complex biological networks. In particular, our aim is to study their performance when applying such quantitative network measures to disease networks (prostate cancer versus benign tissue), i.e., graphs inferred from biological data sets based on different conditions of prostate cancer. The networks we have inferred are from public available micro array studies. Different kinds of topological graph measures (non-information-theoretic and informationtheoretic) are calculated with a subsequently performed cluster analysis. We analyze different sets of descriptors and come to quite reasonable results when clustering between the different conditions of the disease.

• Type: Journal
• Year: 2010
• Where: BIOCOMP
• Authors: Laurin A. J. Mueller, Karl G. Kugler, Andreas Dander, Christian Baumgartner, Armin Graber, Matthias Dehmer
• Keywords: BIOCOMP 2010, Bioinformatics, Information-theoretic Measures, Micro Array Studies, Prostate Cancer