EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Biased sampling driven by bacterial population structure confounds machine learning prediction of antimicrobial resistance

Yanying Yu iD, Nicole E Wheeler iD, Lars Barquist iD

PLOS Biology · 16.12.2025 · 23 (12) · e3003539

DOI: 10.1371/journal.pbio.3003539 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS Biology, online erschienen am 16.12.2025. Verfasst von Yanying Yu und 2 weiteren Autor:innen. Beteiligte Einrichtung: Helmholtz Institute for RNA-based Infection Research (HIRI), Helmholtz Centre for Infection Research (HZI), Würzburg, Germany. Hauptthemen laut MEDLINE: Machine Learning, Drug Resistance, Bacterial, Bacteria. Die Arbeit wurde bisher 26-mal zitiert (Crossref) und nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Bayresq.net, Natural Sciences and Engineering Research Council (NSERC), Bayerisches Staatsministerium für Wissenschaft, Forschung und Kunst.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Antimicrobial resistance (AMR) poses a growing threat to human health. Increasingly, genome sequencing is being applied for the surveillance of bacterial pathogens, producing a wealth of data to train machine learning (ML) applications to predict AMR and identify resistance determinants. However, bacterial populations are highly structured, and sampling is biased towards human disease isolates, violating ML assumptions of independence between samples. This is rarely considered in applications of ML to AMR. Here, we demonstrate the confounding effects of sample structure by analyzing over 24,000 whole genome sequences and AMR phenotypes from five diverse pathogens, using pathological training data where resistance is confounded with phylogeny. We show the resulting ML models perform poorly and that increasing the training sample size fails to rescue performance. A comprehensive analysis of 6,740 models identifies species- and drug-specific effects on model accuracy. These findings highlight the limitations of current ML approaches in the face of realistic sampling biases and underscore the need for population structure-aware methods and more diverse datasets to improve AMR prediction and surveillance.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Machine LearningDrug Resistance, BacterialBacteriaHumansAnti-Bacterial AgentsPhylogenyGenome, BacterialWhole Genome Sequencing

Substanzen

Anti-Bacterial Agents

Autor:innen und Institutionen

  • Helmholtz Institute for RNA-based Infection Research (HIRI), Helmholtz Centre for Infection Research (HZI), Würzburg, Germany.
  • Division of Infectious Diseases, Boston Children's Hospital and Department of Pediatrics, Harvard Medical School, Boston, Massachusetts, United States of America.
  • Institute of Microbiology and Infection, University of Birmingham, Birmingham, United Kingdom.
  • Department of Biology, University of Toronto, Mississauga, Canada.
  • Department of Cell and Systems Biology, University of Toronto, Toronto, Canada.

ORCID-Profile: Yanying Yu, Nicole E Wheeler, Lars Barquist

Förderung

  • Bayresq.net
  • Natural Sciences and Engineering Research Council (NSERC)
  • Bayerisches Staatsministerium für Wissenschaft, Forschung und Kunst
  • NSERC

Zitieren

APA

Yu, Y., Wheeler, N. E., & Barquist, L. (2025). Biased sampling driven by bacterial population structure confounds machine learning prediction of antimicrobial resistance. PLOS Biology, 23(12), e3003539. https://doi.org/10.1371/journal.pbio.3003539

Vancouver

Yu Y, Wheeler NE, Barquist L. Biased sampling driven by bacterial population structure confounds machine learning prediction of antimicrobial resistance. PLOS Biology. 2025;23(12):e3003539. doi:10.1371/journal.pbio.3003539

BibTeX
@article{Yu2025,
  author = {Yu, Yanying and Wheeler, Nicole E and Barquist, Lars},
  title = {Biased sampling driven by bacterial population structure confounds machine learning prediction of antimicrobial resistance},
  journal = {PLOS Biology},
  year = {2025},
  volume = {23},
  number = {12},
  pages = {e3003539},
  doi = {10.1371/journal.pbio.3003539}
}
Kostenlos starten →