Gene network analysis of type 2 diabetes mellitus.

摘要

Type 2 diabetes mellitus is a metabolic disorder that is believed to be caused by a combination of environmental factors and genetic disposition. Despite an impressive body of research and steady progress in understanding associated risk factors such as obesity and insulin resistance, the underlying disease mechanisms remain unknown. In this project, we described and developed a gene network analysis approach, called Gene Network Enrichment Analysis to study of type 2 diabetes mellitus. Gene networks encode relationships between genes; by accounting for such information in addition to gene expression, we show that Gene Network Enrichment Analysis is able to identify biological processes, signaling pathways, and individual genes associated with disease that are missed by standard analyses. We developed multiple variants of the algorithm and show that it is equally suited to analyzing individual experiments or as a meta-analysis across multiple, biologically related experiments. Application of Gene Network Enrichment Analysis on a large compendium of animal models of disease successfully identifies differential activity in insulin signaling, nuclear receptors, metabolism processes, and inflammatory response. Subsequent network analysis comparing diabetes resistant and diabetes prone mouse models demonstrated that differences in inflammation precedes any measurable metabolic differences, and may stem from differences in the actual number of inflammatory cells. Our results therefore suggest that dysregulation in inflammation, specifically T cell and macrophage activation, may be an early feature of disease.;We compare the performance of Gene Network Enrichment Analysis to a number of other tools used in the community including standard hypergeometric enrichment on differentially expressed genes and Gene Set Enrichment Analysis. While we focused on type 2 diabetes mellitus in this project, the Gene Network Enrichment Analysis algorithm that we have developed is generally applicable. To facilitate the adoption of the algorithm by the wider scientific community, we have implemented a publicly available and open-source web server for Gene Network Enrichment Analysis. The server is located at http://soteira.bu.edu .