Rapid Species-Level Classification of Urinary Pathogens from Raw LC-MS/MS Signals Using Machine Learning.
6 affiliations
- 1Computational Biology Laboratory, CHU de Québec-Université Laval Research Center, Québec City, QuebecG1V 4G2, Canada.
- 2Proteomics Platform, CHU de Québec-Université Laval Research Center, Québec City, QuebecG1V 4G2, Canada.
- 3Centre de Recherche en Infectiologie de l'Université Laval, Axe Maladies Infectieuses et Immunitaires, Centre de Recherche du CHU de Québec-UniversitéLaval, Québec City, QuebecG1V 4G2, Canada.
- 4Evosep Biosystems, OdenseDK-5230, Denmark.
- 5ThermoFisher Scientific, Bremen28199, Germany.
- 6Université Côte d'Azur, Inria, CNRS, I3S, Nice06903, France.
Abstract
Urinary tract infections are among the most common infections in humans, yet their diagnosis still depends on time-consuming workflows based on microbial culture, followed by MALDI-TOF mass spectrometry. Although LC-MS/MS offers the sensitivity and specificity needed to bypass culture, conventional pipelines depend on lengthy analyses and peptide/protein identification steps, limiting the throughput and hindering its adoption in clinical settings. Here, we introduce a direct, identification-free LC-MS/MS workflow that analyzes raw ion signal and produces species-level microbial identification in about 5 min after preparation, fast enough to meet clinical throughput requirements. Our machine learning-enabled raw-signal pipeline bypasses peptide identification entirely, preserving information and eliminating the traditional interpretation stack. Across 15 independent analytical batches covering 28 clinically relevant pathogens, the method achieved high-confidence classification (MCC = 0.86). Applied to 206 clinical urine specimens across three batches, the approach reached 91% accuracy at clinically actionable microbial loads (greater than 105 CFU/mL) and, critically, 0 false positives in control specimens. The performance was lower for specimens below this threshold. These results show that raw LC-MS/MS spectra contain sufficient biological information for direct microbial diagnosis, establishing an analytical framework for clinical mass spectrometry. This proof-of-concept demonstrates that rapid, culture-free, fast microbial identification is achievable and positions raw signal inference as a promising direction for next-generation diagnostic mass spectrometry.
Related
- Machine Learning based Classification of Ocular Pathogens using Host-transcriptomic Signatures: a proof-of-concept
- Evaluation of New bioMérieux Chromogenic CPS Media for Detection of Urinary Tract Pathogens
- Component-specific prognostic signals within a multidimensional classification for periprosthetic joint infection and the impact of repeated surgical episodes: a retrospective cohort study
- Machine learning-guided screening of urine cultures for urinary tract infection diagnosis according to AMCLI guidelines
- Machine Learning-Based Fatigue Detection from EEG Signals: A Methodological Pilot Framework with Potential for Nursing Applications
BibTeX
@article{pelletier2026rapid,
title = {{Rapid Species-Level Classification of Urinary Pathogens from Raw LC-MS/MS Signals Using Machine Learning.}},
author = {Simon J Pelletier and Antoine Lacombe-Rastoll and Florence Roux-Dalvai and Mickaël Leclercq and Clarisse Gotti and Eve Bérubé and Pascaline Bories and Marie-Ève Thibeault and Dorte B Bekker-Jensen and Nicolai Bache and Maciej Bromirski and Sandra Isabel and Frederic Precioso and Arnaud Droit},
journal = {Analytical chemistry},
year = {2026},
month = oct,
doi = {10.1021/acs.analchem.6c02725},
url = {https://europepmc.org/article/MED/42915683},
note = {CC BY-NC-ND},
}
RIS
TY - JOUR
TI - Rapid Species-Level Classification of Urinary Pathogens from Raw LC-MS/MS Signals Using Machine Learning.
AU - Simon J Pelletier
AU - Antoine Lacombe-Rastoll
AU - Florence Roux-Dalvai
AU - Mickaël Leclercq
AU - Clarisse Gotti
AU - Eve Bérubé
AU - Pascaline Bories
AU - Marie-Ève Thibeault
AU - Dorte B Bekker-Jensen
AU - Nicolai Bache
AU - Maciej Bromirski
AU - Sandra Isabel
AU - Frederic Precioso
AU - Arnaud Droit
PY - 2026
DA - 2026/10/01
JO - Analytical chemistry
DO - 10.1021/acs.analchem.6c02725
UR - https://europepmc.org/article/MED/42915683
AB - Urinary tract infections are among the most common infections in humans, yet their diagnosis still depends on time-consuming workflows based on microbial culture, followed by MALDI-TOF mass spectrometry. Although LC-MS/MS offers the sensitivity and specificity needed to bypass culture, conventional pipelines depend on lengthy analyses and peptide/protein identification steps, limiting the throughput and hindering its adoption in clinical settings. Here, we introduce a direct, identification-free LC-MS/MS workflow that analyzes raw ion signal and produces species-level microbial identification in about 5 min after preparation, fast enough to meet clinical throughput requirements. Our machine learning-enabled raw-signal pipeline bypasses peptide identification entirely, preserving information and eliminating the traditional interpretation stack. Across 15 independent analytical batches covering 28 clinically relevant pathogens, the method achieved high-confidence classification (MCC = 0.86). Applied to 206 clinical urine specimens across three batches, the approach reached 91% accuracy at clinically actionable microbial loads (greater than 105 CFU/mL) and, critically, 0 false positives in control specimens. The performance was lower for specimens below this threshold. These results show that raw LC-MS/MS spectra contain sufficient biological information for direct microbial diagnosis, establishing an analytical framework for clinical mass spectrometry. This proof-of-concept demonstrates that rapid, culture-free, fast microbial identification is achievable and positions raw signal inference as a promising direction for next-generation diagnostic mass spectrometry.
SN - 0003-2700
SN - 1520-6882
ER -