Abstract 15734: Local Admixture Mapping of Metabolites in African Americans

Abstract

Introduction: There is increasing evidence that metabolite levels in African Americans (AA) are influenced by population differentiated genetic variation. We investigated how local genetic ancestry may be associated with levels of circulating plasma metabolites using metabolomic profiling in the Jackson Heart Study, an AA cohort, in an attempt to identify additional genetic determinants of metabolite levels. Hypothesis: Certain metabolites will be associated with local African genetic ancestry. Methods: We performed admixture mapping using local ancestry estimates (probabilities of whether an individual has 0, 1 or 2 alleles of African ancestry at each site in the genome from RFMix, based on similarity to 1000G reference panel) from TOPMed sequencing data for autosomal SNPs. Associations were then tested for 302 targeted metabolites measured using LC-MS. Each local ancestry regression model was adjusted for age, sex, and estimated global ancestry. We used a previously established significance threshold for local African ancestry analysis, p<2.1X10 -5 . We further adjusted findings for previously reported GWAS variants in conditional analysis. Results: There were 38 local admixture mapping signals for 32 metabolites. Fourteen metabolite signals survived conditional analysis and represent signals distinct from those identified in single variant GWAS. Metabolites of note include the branched chain amino acids, whose association with cardiometabolic disease has been shown to differ based on self-identified race. Conclusions: We present local admixture mapping results providing further evidence that population differentiated genetic variants influence circulating levels of plasma metabolites. For a large proportion of our findings, admixture associations were not attenuated by conditioning on previously reported SNPs from GWAS studies. These analyses suggest that admixture mapping may identify association signals that are missed by GWAS, and may help to elucidate population differences affecting key biologic pathways.