iMet: A computational tool for structural annotation of unknown metabolites from tandem mass spectra

TL;DR

iMet is a computational tool that leverages tandem mass spectrometry data to annotate unknown metabolites by identifying structural similarities with known compounds, enhancing metabolite identification in untargeted studies.

Contribution

This paper introduces iMet, a novel algorithm that uses MS/MS spectra to annotate unknown metabolites by structural comparison, filling a gap in metabolomics analysis.

Findings

89% of metabolites had at least one correct match in top four predictions

Validated on 148 metabolites with high accuracy

Freely available tool for metabolite annotation

Abstract

Untargeted metabolomic studies are revealing large numbers of naturally occurring metabolites that cannot be characterized because their chemical structures and MS/MS spectra are not available in databases. Here we present iMet, a computational tool based on experimental tandem mass spectrometry that could potentially allow the annotation of metabolites not discovered previously. iMet uses MS/MS spectra to identify metabolites structurally similar to an unknown metabolite, and gives a net atomic addition or removal that converts the known metabolite into the unknown one. We validate the algorithm with 148 metabolites, and show that for 89% of them at least one of the top four matches identified by iMet enables the proper annotation of the unknown metabolite. iMet is freely available at http://imet.seeslab.net.