Simulating metabolic pathways to enhance interpretations of metabolome genome-wide association studies

Abstract

Advancements in large-scale analysis of metabolites in human peripheral blood samples revealed the links between metabolite concentrations and genetic variations. This field is known as metabolome-genome-wide association study (MGWAS). Although MGWAS is a powerful tool, it has some limitations, particularly in terms of the number of metabolites that can be measured. Whether the observed associations are directly due to genetic variation or indirectly due to changes in unmeasured metabolites is unclear. To address this, we used simulations of metabolic pathway models to investigate the influence of genetic variants on metabolite concentrations and enhance the interpretation of MGWAS results. By systematically adjusting the enzyme reaction rates to simulate genetic variants, we observed changes in the metabolite levels. Our simulations accurately represented most of the variant-metabolite pairs identified by MGWAS with significant p-values, thereby demonstrating the potential of our approach. Furthermore, our simulations revealed additional marked fluctuations in metabolite levels that the MGWAS did not detect, suggesting that some variant-metabolite pairs might become more significant with larger sample sizes. We also categorized the enzymes into three types based on their impact on metabolite concentrations, highlighting enzymes with minimal impact. This indicated that genetic variations in these enzymes may have limited biological significance. Our study not only validates key MGWAS findings, but also provides a systematic framework for understanding enzyme-metabolite relationships. This approach offers valuable insights for future experimental studies and potential therapeutic interventions.

Keywords: Biological simulation; GWAS; Metabolomics.