Two-dimensional correlation analysis of nuclear magnetic resonance metabonomics data

Area of Science:

• Metabonomics
• Spectroscopy
• Preclinical Safety Assessment

Background:

• Vasculitis induction in rats is a critical preclinical safety evaluation.
• Nuclear magnetic resonance (NMR) metabonomics generates complex spectral data.
• Two-dimensional (2D) correlation analysis offers advanced spectral data interpretation.

Purpose of the Study:

• To apply 2D correlation analysis to NMR metabonomics spectra of rat urine.
• To identify spectral changes associated with induced vasculitis.
• To evaluate the utility of 2D correlation analysis for preclinical safety studies.

Main Methods:

• Collected rat urine samples during induced vasculitis.
• Performed 2D correlation analysis (covariance and correlation coefficient maps) on binned and raw NMR spectra.
• Analyzed spectral data at resolutions of 0.04 and 0.001 ppm.

Main Results:

• Identified good correlation among major peaks in binned spectra.
• Sample-sample 2D correlation maps distinguished two sample groups.
• Raw spectra revealed complex features, including butterfly-like patterns in covariance maps, with fewer significant coefficients in correlation maps.
• Clustering of late-stage samples was observed in sample-sample covariance maps for both binned and raw data.

Conclusions:

• 2D correlation analysis is effective for analyzing NMR metabonomics data in preclinical safety studies.
• Covariance maps, particularly for raw spectra, provide valuable insights into spectral changes.
• The study highlights the importance of spectral data resolution and alignment in 2D correlation analysis.