MetaboVariation: Exploring Individual Variation in Metabolite Levels

Area of Science:

• Metabolomics
• Personalized Medicine
• Biomarker Discovery

Background:

• Current metabolomics research primarily focuses on disease biomarkers.
• There is a critical need for biomarkers indicating early metabolic dysfunction to guide lifestyle interventions.
• Analyzing metabolomics data at the individual level requires novel strategies.

Purpose of the Study:

• To introduce MetaboVariation, a novel method for analyzing metabolite level fluctuations in individuals.
• To enable the identification of individuals with early metabolic dysfunction using repeated metabolomics measurements.
• To facilitate personalized healthcare through individual-level metabolomics data analysis.

Main Methods:

• MetaboVariation utilizes a Bayesian generalized linear model to assess intra-individual metabolite variations.
• The method models repeated metabolite levels, accounting for individual-specific fluctuations and explanatory variables.
• Individuals are flagged if their observed metabolite levels fall outside the 95% highest posterior density prediction interval.

Main Results:

• MetaboVariation was applied to metabolomics data from 164 individuals with repeated measurements.
• A significant percentage of individuals (28%) were identified with intra-individual variations in three or more metabolites.
• An R package with an R Shiny web application was developed for MetaboVariation.

Conclusions:

• MetaboVariation represents a significant advancement in analyzing metabolomics data at the individual level.
• The method aids in identifying early metabolic dysfunction, crucial for timely lifestyle interventions.
• This approach paves the way for more personalized healthcare strategies based on individual metabolic profiles.