Recoupled-STOCSY-based co-expression network analysis to extract phenotype-driven metabolite modules in NMR-based...

Area of Science:

• Metabolomics
• Systems Biology
• Biomedical Data Analysis

Background:

• Traditional Nuclear Magnetic Resonance (NMR)-based metabolomics faces limitations in spectral preprocessing, biomarker identification, and biological interpretation.
• Existing methods struggle with inherent defects in spectral alignment and dimensionality reduction.

Purpose of the Study:

• To propose a novel analytical framework combining Recoupled Statistical Total Correlation Spectroscopy (RSTOCSY) and Weighted Gene Co-expression Network Analysis (WGCNA).
• To enhance spectral alignment, dimensionality reduction, differential metabolite screening, and biological explanation of metabolic networks.
• To evaluate the performance of the RSTOCSY-WGCNA method using serum NMR data from coronary heart disease with diabetes mellitus (CHDDM) patients.

Main Methods:

• Statistical Recoupling of Variables (SRV) was employed for spectral alignment and dataset categorization into signal superclusters.
• Weighted Gene Co-expression Network Analysis (WGCNA) was used to identify phenotype-driven metabolite modules.
• Metabolite identification and correlation analysis were performed within and between identified modules using 2D pseudospectrum.

Main Results:

• The RSTOCSY-WGCNA method demonstrated effectiveness in spectral alignment, validated on a wine dataset.
• Three metabolite modules associated with CHDDM were identified, leading to the identification of 22 key metabolites.
• Analysis revealed significant metabolic correlations and implicated the ferroptosis pathway in CHDDM pathogenesis, involving amino acid, microbial, and glucose metabolism.

Conclusions:

• The proposed RSTOCSY-based WGCNA approach offers an effective workflow for metabolite information recovery and structure identification.
• This method significantly improves the interpretability and understanding of disease pathogenesis.
• The study highlights a novel link between the ferroptosis pathway and CHDDM.