PNNL

Abstract

In the current study we present an approach to identify potential mediators of important immune response processes; combining a network topology approach with comparative analysis of differentially expressed genes shared in multiple stimuli responses. We first show that highly central genes in co-expression networks inferred from transcriptional data are significantly enriched in genes with protein products that are known targets of pathogens; indicating their importance to the system. We identify a common core response module that is shared under all conditions examined and show that this module is significantly enriched in conserved genes and genes with protein products known to be targets of pathogens. Finally, we use a machine-learning method to develop a highly accurate predictive model of the core response module, and identify known and novel regulators of sets of these genes. This study provides a number of interesting and novel insights into macrophage response to pathogens, and outlines a valuable and informative approach to identify critical nodes in the host response to pathogens.