MSclassifR: an R package for supervised classification of mass spectra with machine learning methods

Alexandre Godmer (1, 2) , Yahia Benzerara (2) , Nicolas Veziris (1, 2) , Mariette Matondo (3) , Alexandra Aubry (1, 4) , Quentin Giai Gianetto (3, 5)

1 CIMI - Centre d'Immunologie et des Maladies Infectieuses
2 CHU Saint-Antoine [AP-HP]
3 Plateforme de Protéomique / Proteomics platform
4 CHU Pitié-Salpêtrière [AP-HP]
5 Hub Bioinformatique et Biostatistique - Bioinformatics and Biostatistics HUB

Motivation: Classification of mass spectra is essential for identifying microorganisms from matrix-assisted laser desorption/ionization-time of flight (MALDI-TOF) mass spectrometry. However, spectrally close organisms remain difficult to identify. In this context, we developed the MSclassifR R package to improve the classification of mass spectra. Its open code strengthens the reproducibility of analyzes in the community. Results: We applied the functions of our package to raw mass spectra from three different laboratories. The best workflow available in MSclassifR package achieves near 100% accuracy in all three datasets. Thus, MSclassifR constitutes an interesting alternative for reliable MALDI-TOF based diagnosis. Availability: MSclassifR is freely available online from CRAN repository https://cran.r-project.org/web/packages/MSclassifR/index.html . Two vignettes illustrating how to use the functions of this package from real data sets are also available online to help users.

Domaines: Sciences du Vivant [q-bio]; Statistiques [stat]

Alexandre Godmer, Yahia Benzerara, Nicolas Veziris, Mariette Matondo, Alexandra Aubry, et al.. MSclassifR: an R package for supervised classification of mass spectra with machine learning methods. 2023. ⟨pasteur-04093599v1⟩