Autonomous METLIN-Guided In-source Fragment Annotation for Untargeted Metabolomics

Area of Science:

• Metabolomics
• Mass Spectrometry
• Computational Biology

Background:

• Untargeted metabolomics aims to identify metabolites but often struggles with unannotated features.
• Current annotation methods primarily rely on adduct detection, leaving many features unidentified.
• In-source fragments, common in liquid chromatography-electrospray ionization-mass spectrometry, are underutilized for metabolite annotation.

Purpose of the Study:

• To develop and evaluate a novel strategy for annotating in-source fragments in untargeted metabolomics data.
• To improve metabolite identification by utilizing low-energy tandem mass spectrometry (MS) spectra.
• To enhance the comprehensiveness and confidence of metabolite annotation in complex biological samples.

Main Methods:

• Introduction of MISA (METLIN-guided in-source annotation) algorithm.
• Comparison of detected features against low-energy MS/MS spectra from the METLIN library.
• Evaluation using liquid chromatography-mass spectrometry data from 140 metabolites across three sample sets.

Main Results:

• MISA successfully identified neutral molecular masses using in-source fragments when adducts were not detected.
• The algorithm provided putative metabolite identities based on matched fragments and spectral intensity similarity.
• In-source fragmentation was confirmed as a frequent phenomenon crucial for comprehensive annotation.

Conclusions:

• In-source fragment annotation offers a complementary approach to adduct-based methods, increasing identification confidence.
• The MISA strategy significantly enhances the ability to annotate and identify metabolic features in untargeted profiling.
• MISA is integrated into the XCMS Online platform, providing a freely accessible tool for the metabolomics community.