Multivariate modeling strategy for intercompartmental analysis of tissue and plasma 1H NMR spectrotypes

Area of Science:

• Metabolomics
• Chemometrics
• Systems Biology

Background:

• Understanding metabolic relationships between biological compartments is crucial for systems biology.
• Multicompartmental metabolic profiling offers insights into complex physiological processes.
• Existing methods have limitations in revealing intercompartmental functional links.

Purpose of the Study:

• To apply unsupervised chemometric methods for integrating metabolic profiles from multiple mouse biological matrices.
• To infer functional links between different compartments (plasma, liver, pancreas, adrenal gland, kidney cortex).
• To characterize compartment-specific metabolite signatures (spectrotypes) and assess their contribution.

Main Methods:

• 1H NMR metabolic profiling of plasma, liver, pancreas, adrenal gland, and kidney cortex.
• Application of Principal Component Analysis (PCA) and Multiway PCA.
• Integration of metabolic profiles using Multivariate Curve Resolution (MCR) and Parallel Factor Analysis (PARAFAC).

Main Results:

• PCA showed metabolic differences between matrices but limited intercompartment relationships.
• Multiway PCA assessed interindividual variability and correlations across compartments.
• MCR and PARAFAC provided functional relationships and characterized compartment-specific spectrotypes.
• MCR-ALS and PARAFAC proved effective for variable and compartment selection.

Conclusions:

• MCR and PARAFAC are well-suited for analyzing multicompartmental metabolic data.
• Characterizing spectrotypes enhances understanding of organ-specific metabolic signatures.
• These chemometric approaches offer new avenues for assessing drug or nutritional interventions.