Genome-wide association studies (GWAS) have identified many genetic loci associated with Alzheimer's disease (AD). However, interpretation of these associations remains challenging, in part due to a lack of consideration of molecular mechanisms linking the genome and AD phenome in existing analytic approaches. In addition, with the failure of numerous drug trials, it is of great interest to identify novel, causal, and targetable risk factors for AD. Metabolomic changes in plasma and cerebrospinal fluid (CSF) may be robust biomarkers for AD. If causality can be demonstrated, they also have the potential to be therapeutic targets. Further, systematic integration of metabolomic data with GWAS associations may provide mechanistic insights into the genetic basis of AD and further our understanding of the genetic variants associated with AD outcomes. Here, we propose a set of projects that will integrate highly original genetic and multi-tissue metabolomic data from the Wisconsin Registry of Alzheimer's Prevention (WRAP) and Wisconsin Alzheimer's Disease Research Center (W-ADRC) with summary data from large-scale AD GWAS conducted by consortia to identify novel metabolomic risk factors for AD. The overarching goal of this study is to further scientific understanding of the genetic regulation of metabolome, identify robust metabolite-AD associations in plasma and CSF, and estimate the causal effects of metabolomic variations on AD outcomes. Innovations in analytic methods will manifest in novel statistical approaches to conduct metabolome-wide association studies using GWAS summary statistics and robust Mendelian randomization approaches for causal inference. Leveraging these advanced statistical methods developed by the investigators, our proposed study will enhance the statistical power by using summary statistics from large-scale AD GWAS, reduce the impact of confounding and reverse causality by utilizing genetic prediction models for metabolites built from dementia-free reference cohorts, and assess metabolite-AD associations and possible causal effects through hypothesis-free scans. These advances in data and statistical methodology provide a unique opportunity to identify metabolomic risk factors that are both statistically and clinically meaningful for AD. These results will reveal fundamental new insights into AD etiology and provide analytic tools that are widely applicable in human genetics research.
This proposal includes a set of projects that will integrate Alzheimer's genetic associations and metabolomic data from plasma and cerebrospinal fluid tissues to identify metabolomic risk factors for Alzheimer's disease. The projects will apply novel statistical methods developed by the investigators to advance our understanding of the genetic regulation of metabolome, conduct metabolome-wide association studies for Alzheimer's disease, and estimate potential causal effects. The association and causal results will shed light on the understudied metabolomic pathways linking the genome and AD phenome.
