Metabolic reaction network-based recursive metabolite annotation for untargeted metabolomics

Area of Science:

• Metabolomics
• Biochemistry
• Bioinformatics

Background:

• Untargeted metabolomics using liquid chromatography-mass spectrometry (LC-MS) faces challenges in large-scale metabolite annotation.
• Accurate identification of metabolites is crucial for understanding biological systems and disease mechanisms.

Purpose of the Study:

• To develop a novel algorithm, MetDNA, for expanding metabolite annotations in LC-MS-based untargeted metabolomics.
• To overcome the limitations of requiring comprehensive standard spectral libraries for metabolite identification.

Main Methods:

• Development of a metabolic reaction network (MRN)-based recursive algorithm named MetDNA.
• Utilizing seed metabolites and their reaction-paired neighbors, leveraging MS2 spectral similarities.
• Employing experimental MS2 spectra of known metabolites as surrogates for annotating unknown neighbors.

Main Results:

• MetDNA successfully expands metabolite annotations without a comprehensive spectral library.
• Demonstrated utility and versatility across different LC-MS platforms, acquisition methods, and biological samples.
• Achieved cumulative annotation of approximately 2000 metabolites from a single experiment.

Conclusions:

• MetDNA significantly enhances metabolite annotation capabilities in untargeted metabolomics.
• The algorithm facilitates quantitative assessment of metabolic pathways.
• Enables more robust integrative multi-omics analyses by improving metabolite identification.