Multi-omics characterization of early chronic obstructive pulmonary disease

Area of Science:

• Pulmonary Medicine
• Biomarker Discovery
• Multi-omics Research

Background:

• Chronic obstructive pulmonary disease (COPD) is a leading global cause of death, with early diagnosis crucial for management.
• Current early COPD (ECOPD) diagnosis relies on lung function and smoking history, which may not fully reflect disease progression.
• GOLD guidelines advocate for biological markers over clinical symptoms for improved early detection.

Purpose of the Study:

• To explore the biological characteristics of ECOPD using a multi-omics approach.
• To identify plasma protein and metabolite signatures associated with ECOPD.
• To investigate the potential of multi-omics data for ECOPD subgrouping and lung function prediction.

Main Methods:

• Proteomics and metabolomics analysis of plasma samples from 88 ECOPD patients and 88 healthy controls.
• Univariable logistic regression and gene set enrichment analysis to identify significant proteins.
• Machine learning models and similarity network fusion for subgroup analysis and prediction.

Main Results:

• Proteomics identified 248 proteins associated with ECOPD, predominantly involved in inflammation.
• Metabolomics linked 137 metabolites to ECOPD.
• A multi-omics approach best predicted lung function (R²=0.74), while proteomics alone diagnosed ECOPD (AUC=0.949).
• Two ECOPD subgroups were identified: one driven by inflammatory-immune responses, the other by hemostasis/vascular smooth muscle markers.

Conclusions:

• Multi-omics integration offers a powerful strategy for distinguishing ECOPD subgroups.
• Identified protein and metabolite signatures hold promise for early ECOPD diagnosis and risk stratification.
• This research advances understanding of ECOPD heterogeneity and potential therapeutic targets.