Correlation guided Network Integration (CoNI) reveals novel genes affecting hepatic metabolism

Obesity-induced ectopic fat disposition in the liver is a major risk factor in the pathogenesis of type 2 diabetes as it impairs hepatic insulin sensitivity, a crucial component of whole body glucose homeostasis; however molecular mechanisms remain largely elusive. Understanding the pathogenesis of fatty liver, including the identification of novel genetic regulators that are involved in the regulation of liver metabolism under excess energy supply, is crucial for the development and implementation of efficient prevention and treatment strategies. We here developed a new method to integrate metabolomics and transcriptomics data based on pairwise correlation analysis of metabolites coupled to partial correlation combining the metabolite correlations with the transcript expression profiles followed by the construction of undirected, weighted graphs.This Correlation based Network Integration (CoNI) approach was applied to liver metabolome and transcriptome datasets of lean and HFD-fed obese mice to unravel previously hidden local regulator genes (LRG). The selected candidate genes were validated by transcriptome-proteome correlation analysis, by association studies with liver lipid metabolism in humans and by analysis of cellular metabolite levels after siRNA knockdown. Overall, the new bioinformatic CoNI approach for Omics datasets allowed us to identify genes regulating metabolic networks in livers of obese mice that if solely analyzing the transcriptome dataset would have remained hidden. We performed gene expression microarray analysis of liver from mice treated with high fat diet and control mice