NIAGADS Metalbolomics

We propose to analyze and replicate MGWAS (Metabolome-Wide Association Studies) findings using blood samples from several population-based cohorts. This dataset will serve to replicate metabolite-SNP association signals from blood from several population-based cohorts.
Additionally, we plan to conduct Mendelian Randomization (MR) and mediation analyses for pleiotropic SNPs (single nucleotide polymorphisms) identified in our study. Finally, we will utilize this dataset to test and fine-tune a multiomics machine learning prediction model for Alzheimer's Disease (AD).
Specifically, this metabolomics dataset, obtained from both brain and cerebrospinal fluid (CSF), will be instrumental in validating our findings from blood metabolomic datasets. This validation step is crucial for establishing a systematic understanding of the shared genetic basis of metabolites and AD, potentially leading to early prediction and intervention strategies for AD.
By integrating data from multiple biological sources, we aim to enhance the robustness and reliability of our findings. This dataset will serve as an additional validation resource, providing a solid foundation for advancing our knowledge of AD's genetic underpinnings and developing predictive models.

Metabolites (i.e. quantified small molecules) can reflect what has been encoded by the genome and modified by environmental factors at a specific point in time. Such biomarkers could be particularly useful for diagnosis improvement, potential therapeutic targets, as well as additional insights into disease pathogenesis. We hypothesize that many metabolites are strong intermediate variables (mediators) of genetic effects and/or environmental effects on AD. Specifically, several metabolite quantitative trait loci (metQTLs) affect biomarkers across the human life course starting from childhood and indirectly leading to AD, while others are influenced by lifestyle and environment later in life. To understand AD mechanisms, we will systematically study SNP-metabolite associations in blood and integrate them with AD susceptibility loci. We will use the NG00131 cohort to validate our findings in metabolites from brain and CSF, as the tissues of interest for AD.