Dissertations & Theses (Open Access)

Date of Award


Degree Name

Master of Public Health (MPH)


Han Chen

Second Advisor

Paul De Vries


The aim of the study was to identify associations between predicted gene expression levels using S-PrediXcan and carotid intima media thickness (cIMT) in order to elucidate the etiology of atherosclerosis. Further, we wanted to compare our results to those from a transcriptome wide association study (TWAS) that used actual gene expression levels in whole blood. For our analysis we used summary statistics from a GWAS of cIMT from the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) consortium, tissue-specific gene expression profiles from the GTEx project, and S-PrediXcan prediction models. The TWAS of cIMT that we compared our results to used gene expression levels in peripheral whole blood of 5,647 individuals from the CHARGE consortium. After quality control measures and limiting our results to a single tissue-gene pair per locus in arterial tissues, we identified 4 novel loci that had not previously been associated with cIMT including locus near RP11-227D13.1, CCDC34, CD52, and RAB6A. When restricting analysis to whole blood, correlation between predicted gene expression levels using S-PrediXcan and actual gene expression using TWAS had low correlation (Pearson’s r 0.0241).