Improving lipid mapping in Genome Scale Metabolic Networks using ontologies

Area of Science:

• Metabolomics and Systems Biology
• Bioinformatics and Computational Biology

Background:

• Interpreting metabolomic and lipidomic data requires connecting measured molecules within metabolic networks.
• Challenges exist in mapping experimental data to metabolic networks due to identifier and annotation discrepancies, particularly for lipids.

Purpose of the Study:

• To develop a flexible method for linking lipids from lipidomics datasets to metabolic networks.
• To improve the integration of experimental metabolomic data with genome-scale metabolic network reconstructions.

Main Methods:

• A novel matching method utilizing the ChEBI ontology was developed.
• The method computes distances between molecules in the ChEBI ontology to establish flexible links.
• The implementation is available as a Python library and within the MetExplore webserver.

Main Results:

• The ontology-based mapping demonstrated improved and facilitated linking of lipids to genome-scale metabolic networks.
• Application to a lipid library and experimental dataset confirmed the method's efficacy.
• The approach offers improvements for metabolic network curation and lipidomics data annotation.

Conclusions:

• The generic, ontology-based method enhances the comprehension of metabolic modulations by enabling robust data integration.
• This approach is applicable to diverse metabolomics datasets, advancing systems biology research.