Characterization of disease associated genes in high confidence protein- protein interaction networks.

ARIEL CHERNOMORETZ

Villa Carlos Paz

Workshop; Latin American Workshop in Non-Linear Physics; 2013

In recent years there has been a lot of interest in the application of complex network theoryto human health related research programs. Taking advantage of the increasing availabilityof molecular interaction data, the network metaphor has appeared as an appealing frameworkto unveil patterns of biological relevance. In order to elucidate the molecular basis of humandiseases a lot of attention has been paid to the analysis of potential biases in topological features- such us degree, clustering or betweenness - for diseases associated genes in the context ofprotein interaction networks. Following this line of research, we present a novel analysis basedon the study of non-trivial associations between disease genes and the modular structure ofthe considered protein interaction network. We found that topological roles played by disease-associated genes were selectively enriched in dierent class of diseases. Our analysis also shows that cancer associated genes tend to populate high participation roles and recapitulates the ideathat cancer genes present dierent topological features than genes associated to other diseaseclasses.