Immune and inflammatory resolution pathways through multi-omics using the AI-based Network Integration

Area of Science:

• Immunology
• Systems Biology
• Genomics
• Metabolomics
• Glycomics

Background:

• Multi-omics data integration is challenging due to data heterogeneity, high dimensionality, and missing values.
• Understanding immune regulation and inflammation resolution requires integrating diverse molecular profiles.
• Existing methods struggle to model complex inter-omics relationships effectively.

Purpose of the Study:

• To propose and validate an AI-based Network Integration approach for multi-omics data.
• To investigate the relationship between immunoglobulin G N-glycans (glycans) and specialized pro-resolving lipid mediators (SPMs) in immune and inflammatory pathways.
• To identify molecular causal networks and bridging molecules influencing immune regulation.

Main Methods:

• Cross-sectional multi-omics study (genetic, glycomic, lipidomic data).
• AI-based Network Integration for modeling inter-omics connectivity.
• Pairwise association analysis and conditional analysis.
• Principal Component Analysis (PCA) for polygenic factor derivation.
• Granularity Directed Acyclic Graph (G-DAG) algorithm for Mendelian randomization.

Main Results:

• AI-based Network Integration successfully modeled complex biological system influences across omics layers.
• Identified 10 significant glycan-SPM associations, with two bridging molecules (5-HEPE SPM and GP21 glycan) mediating these links.
• Discovered that both 5-HEPE and GP21 are under polygenic control, with distinct roles as receivers and broadcasters, respectively.
• Revealed a biological link between SPM lipidomics variations and glycosylation changes.

Conclusions:

• The AI-based Network Integration approach provides interpretable and biologically meaningful insights from noisy multi-omics data.
• Findings suggest a novel connection between lipid mediators and glycan-mediated immune regulation and inflammation resolution.
• The study highlights the utility of integrating genetic, glycomic, and lipidomic data for uncovering complex biological pathways.