Modeling and Classification of Kinetic Patterns of Dynamic Metabolic Biomarkers in Physical Activity

Area of Science:

• Metabolomics and Systems Biology
• Exercise Physiology
• Biomarker Discovery

Background:

• Understanding human metabolic responses to external stimuli like physical activity is crucial for health and disease research.
• Dynamic metabolic biomarkers can provide insights into real-time physiological changes.
• Kinetic modeling offers a powerful approach to characterize these dynamic responses.

Purpose of the Study:

• To classify dynamic metabolic biomarker candidates in human metabolism.
• To model and characterize the kinetic regulatory mechanisms in response to physical activity.
• To identify characteristic kinetic signatures and response patterns.

Main Methods:

• Utilized longitudinal metabolic concentration data from a cycle ergometry cohort study.
• Employed targeted metabolomics with tandem mass spectrometry (MS/MS) and stable isotope dilution (SID) for quantitation.
• Applied mathematical modeling (polynomial fitting) and hierarchical cluster analysis for kinetic signature identification and classification.

Main Results:

• Identified and classified dynamic metabolic biomarker candidates based on maximum fold changes (MFCs) and statistical significance.
• Characterized kinetic response patterns, including sustained, early, and late responses.
• Acetylcarnitine (C2) and alanine were identified as strong predictors with late response patterns; glucose showed a delayed response as a moderate predictor.

Conclusions:

• The developed modeling approach is effective for dynamic biomarker identification in longitudinal studies.
• Kinetic signatures provide valuable insights into metabolic regulation during physical activity.
• This methodology holds potential for biomarker discovery in disease and pharmacodynamics studies.