A Map of the Lipid-Metabolite-Protein Network to Aid Multi-Omics Integration

Area of Science:

• Computational Biology
• Systems Biology
• Biochemistry

Background:

• Understanding complex diseases requires integrating multi-omics data.
• Existing methods often struggle to unify diverse molecular information like lipids, metabolites, and proteins.
• A unified framework is needed to explore cross-omics interactions and biological mechanisms.

Purpose of the Study:

• To develop a unified framework for multi-omics data integration using a lipid-metabolite-protein network.
• To visualize and analyze molecular connections across different omics layers.
• To identify novel biomarkers and understand disease mechanisms and therapeutic effects.

Main Methods:

• Developed a network integrating protein-protein interactions with metabolite and lipid data.
• Utilized hyperbolic embedding for network visualization and analysis.
• Applied a user-friendly Shiny R software package for intuitive exploration.
• Performed functional enrichment analysis to identify biological processes.

Main Results:

• Identified known and potential novel biomarkers for cardiovascular disease (CVD).
• Confirmed associations of cholesterol esters and sphingomyelin with CVD-related proteins.
• Highlighted 4-imidazoleacetate and indoleacetaldehyde as potential novel biomarkers.
• Analyzed temporal effects of empagliflozin on lipid metabolism, revealing dynamic shifts in phospholipid and sphingolipid pathways.

Conclusions:

• The developed multi-omics network provides a versatile tool for hypothesis generation, biomarker discovery, and functional analysis.
• This framework enhances the understanding of disease mechanisms and drug effects by bridging molecular layers.
• The approach has broad applications in computational biology, precision medicine, and drug discovery.