Source: Masters Abstracts International, Volume: 57-05.

Adviser: Andrew DeWan.

INTRODUCTION: Bloodstream infections (BSI) cause over 600 deaths per day in the United States. The most common treatment is broad-spectrum antibiotics which are not cost-effective. These can fail to target the infecting microbe, can contribute to antibiotic resistance, and can cause toxicity in some patients. Therefore new strategies are required to combat BSIs. A previous genome-wide association study (GWAS) has identified single nucleotide polymorphisms (SNPs) which are associated with risk of BSIs. Genetic risk scores (GRS) can be developed from these GWASs.

OBJECTIVES: To develop a GRS to evaluate the cumulative effect of multiple SNPs on BSIs.

METHODS: Using available GWAS data, we applied a GRS analysis to identify if there were groups of SNPs that together predicted risk for BSIs. The GRS was formed in one population, HUNT, and evaluated in another, Tromso.

RESULTS: The GRSs were significantly different between cases and controls for nearly every group of SNPs in the HUNT population. The GRS difference was non-significant in the Tromso population except for rs226985, which was significantly associated with gram positive infections (p=0.0192), and rs3778630, which was significantly associated with gram negative infections (p=0.0411).

CONCLUSION: The GRSs developed are not a valid predictor for BSIs. While the top SNPs for each infection type were significant, there is not an additive effect for risk of BSI when an individual has more than one SNP associated with BSI.