Comparison of Co-Expression Measures: Mutual Information, Correlation, and Model-Based Indices

• Activity Number: 663
• Type: Invited
• Date/Time: Thursday, August 8, 2013 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Steve Horvath*+ and Lin Song
• Companies: University of California, Los Angeles and University of California, Los Angeles
• Keywords: co-expression ; network ; mutual information ; correlation network ; clustering ; dependence measure

Abstract

Mutual information (MI) is often used as a generalized correlation measure. Given that the calculation of MI complex it is important to determine how much MI adds beyond standard (robust) correlation measures or regression model based association measures. Here we provide a comprehensive empirical comparison between mutual information and several correlation measures. We also study different approaches for transforming an adjacency matrix, e.g. using the topological overlap measure. Overall, we confirm close relationships between MI and correlation in all data sets which reflects the fact that most gene pairs satisfy linear relationships. The biweight midcorrelation outperforms MI in terms of elucidating gene pairwise relationships. Coupled with the topological overlap matrix transformation, it often leads to more significantly enriched co-expression modules. Spline and polynomial networks form attractive alternatives to MI in case of non-linear relationships.