 Resistant hypertension is persistently elevated blood pressure, despite concurrent use of multiple anti-partisan medications. It substantially increases the risk of heart attack, congestive heart failure, chronic kidney disease and stroke. African Americans are disproportionately affected by resistant hypertension, as well as its early onset compared to other ethnicities in the United States. With the aim of identifying a candidate gene for treating resistant hypertension, I built a predictive network model, using data I gathered from several African American patients with and without this disease. In this study, I used three classes of variables, clinical variables, expression levels of 434 genes previously associated with blood pressure regulation and literature, as well as 707 additional genes extracted from our data set based on their correlation with the hypertensive state. The model that I built is composed of several decision trees. In each decision tree, multiple variables collective decide the outcome. Each tree branch is split into two branches based on a variable's threshold. Using data from two-thirds of the samples, these thresholds are optimized by creating the optimal separation between healthy and diseased patients. With the remaining one-third of the samples, model predictions are generated using a democratic voting scheme. If more trees vote for the healthy state, a patient is predicted to be disease-free. For such models to be effective in personalized medicine efforts, for saving lives, their accuracy is the key. When I only used clinical variables, the predictive accuracy was 87 percent. With the addition of gene expression levels, the accuracy went up to 91 percent. Next, I used the same approach, this time for predicting the expression level of each gene in terms of all the remaining genes, and identify the gene-gene interaction network. Within this network, a gene named CIF-5R4 turned out to be the gene with the maximum number of connections. CIF-5R4 is emolved in multiple stress-response pathways that are highly active in patients with metabolic disorders. Through mass model studies, it has been linked to mass of fat tissue, rate of fatty acid oxidation, high blood lipid levels, and diabetes, which are all important drivers of vascular pathology. My findings suggest that by perturbing the activity of CIF-5R4, we can prevent some of the adverse effects of resistant hypertension, including organ damage, especially among people who are genetically susceptible. This would help us get closer to a major aim in the field of cardiovascular disease that is preventing deaths caused by hypertension.