Abstract 16106: Identification of Disrupted Network Modules and Key Regulators of Coronary Artery Disease via Integration of Blood Gene Expression Profiling and Network Analysis
Genome-wide association studies (GWAS) and gene expression profiling experiments have successfully identified genes associated with coronary artery disease (CAD) and its risk factors. However, the molecular mechanisms underlying these gene-disease associations remain unclear. We hypothesized that genetic variants with both strong and subtle effects drive gene subnetworks that in turn affect risk of CAD. We surveyed CAD-associated molecular interactions by constructing co-expression networks using gene expression profiles of whole blood samples from 189 prevalent CAD cases and 189 age- and sex-matched controls from Framingham Heart Study (FHS) participants. The CAD cases included 3 subtypes: myocardial infarction (MI, n=98), coronary artery bypass grafting (CABG, n=43), and percutaneous transluminal coronary angioplasty with or without a stent (PTCA, n=48). A total of 24 coexpression network modules were identified from the CAD subtypes and controls. Based on module overlap analysis, two of these were differential modules (DMs), one case-specific and one control-specific. The DMs were enriched for genes involved in B-cell activation, immune response and transporter / channel activity. By integrating the CAD DMs with SNPs associated with altered gene expression (eSNPs) and GWAS of CAD, blood pressure, lipid traits, body mass index (BMI), and diabetes, the control-specific DM was implicated as a putative CAD-causal DM based on its significant enrichment for genes whose eSNPs showed low p value associations with both CAD phenotypes and lipid traits. This putative CAD-causal DM was further integrated with tissue-specific Bayesian networks and protein-protein interaction (PPI) networks to identify regulatory key driver (KD) genes. Multiple tissue/network-specific KDs were identified (CD79B and CTGF for blood, SPIB and BCL2A1B for kidney, and TNFRSF13C and PIK3AP1 for adipose tissue). Our network-driven integrative analysis not only identified CAD subtype-specific network modules and genes, but it also defined a network structure that sheds light on the molecular interactions of CAD risk genes within, between, and across tissues.
- Coronary artery disease
- Gene expression
- Systems biology
- Genome-wide association studies (GWAS)
- © 2012 by American Heart Association, Inc.