Analysis and classification of 1H-NMR spectra by multifractal analysis

Area of Science:

• Metabolomics
• Statistical analysis
• Biophysics

Background:

• Traditional methods for analyzing metabolic processes using specific biomarkers are limited.
• A novel statistical methodology is needed to assess the impact of sulfur amino acids (SAA) on metabolic variability.
• Proton nuclear magnetic resonance (1H-NMR) spectroscopy offers a comprehensive view of metabolic profiles.

Purpose of the Study:

• To develop and validate a statistical methodology for evaluating the effect of SAA intake on metabolic processes.
• To utilize multifractal analysis of 1H-NMR spectra to measure metabolic inhomogeneity.
• To differentiate metabolic patterns associated with varying SAA levels and depletion/repletion states.

Main Methods:

• Multifractal analysis of 1H-NMR spectra using wavelet-based multifractal spectrum.
• Application of two statistical models (Model-I and Model-II) analyzing spectral mode, left slope, and broadness.
• Discriminatory analysis using support vector machine (SVM) and principal component analysis (PCA).

Main Results:

• SAA group effect (high vs. low doses) was significant in both models.
• Model-II showed significant time and depletion/repletion effects for spectral mode.
• 1H-NMR spectra from SAA low groups exhibited more regularity and variability than SAA high groups.
• High and low SAA groups were easily distinguishable; depletion/repletion was distinguishable in both models.

Conclusions:

• Multifractal analysis of 1H-NMR spectra is a powerful tool for metabolic investigation.
• SAA intake significantly influences the hourly variation of metabolic processes and daily depletion/repletion differences.
• The developed methodology provides a more informative approach compared to traditional biomarker analysis.