MARS: A Multipurpose Software for Untargeted LC-MS-Based Metabolomics and Exposomics.
Laura GoracciPaolo TiberiStefano Di BonaStefano BonciarelliGiovanna Ilaria PasseriMarta PiroddiSimone MorettiClaudia VolpiIsmael ZamoraGabriele CrucianiPublished in: Analytical chemistry (2024)
Untargeted metabolomics is a growing field, in which recent advances in high-resolution mass spectrometry coupled with liquid chromatography (LC-MS) have facilitated untargeted approaches as a result of improvements in sensitivity, mass accuracy, and resolving power. However, a very large amount of data are generated. Consequently, using computational tools is now mandatory for the in-depth analysis of untargeted metabolomics data. This article describes MetAbolomics ReSearch (MARS), an all-in-one vendor-agnostic graphical user interface-based software applying LC-MS analysis to untargeted metabolomics. All of the analytical steps are described (from instrument data conversion and processing to statistical analysis, annotation/identification, quantification, and preliminary biological interpretation), and tools developed to improve annotation accuracy (e.g., multiple adducts and in-source fragmentation detection, trends across samples, and the MS/MS validator) are highlighted. In addition, MARS allows in-house building of reference databases, to bypass the limits of freely available MS/MS spectra collections. Focusing on the flexibility of the software and its user-friendliness, which are two important features in multipurpose software, MARS could provide new perspectives in untargeted metabolomics data analysis.
Keyphrases
- mass spectrometry
- liquid chromatography
- high resolution mass spectrometry
- data analysis
- ultra high performance liquid chromatography
- tandem mass spectrometry
- gas chromatography
- high performance liquid chromatography
- ms ms
- simultaneous determination
- solid phase extraction
- high resolution
- big data
- electronic health record
- gas chromatography mass spectrometry
- optical coherence tomography
- machine learning
- deep learning