2008 George L. Duff Memorial Lecture—Cardiovascular Disease: From Genetics to Systems Biology
Cardiovascular disease involves hundreds of genes, many environmental factors, and myriad interactions. Although studies of Mendelian traits and transgenic mice have been very informative, analyzing the individual components of a system is not sufficient. We would like to know how these components interact with each other and how these interactions differ in disease states. Toward this end, we have over the last several years attempted to develop an approach that integrates common DNA variations, global gene expression array analyses, and clinical phenotypes. The approach uses segregating populations of mice or randomized populations of human subjects that are typed for genetic markers as well as clinically relevant traits. In addition, we carry out global expression array analyses of tissues or cells from the subjects. This approach greatly facilities gene discovery, since one can examine correlations between transcript levels and clinical traits. It also permits the construction of gene coexpression networks that underlie cardiovascular disease.
We have applied the approach in different ways to examine clinically relevant traits. First, we used knowledge of loci controlling transcript levels to help identify genes for vascular calcification and high-density lipoprotein levels. Second, we used probabilistic modeling to predict several novel genes involved in metabolic syndrome. Third, we used primary endothelial and smooth muscle cell lines from human subjects to construct gene networks for inflammatory responses.