Automated web-based typing of Clostridioides difficile ribotypes via MALDI-TOF MS

Area of Science:

• Microbiology
• Infectious Diseases
• Computational Biology

Background:

• Clostridioides difficile causes significant hospital-acquired infections and outbreaks.
• Accurate and rapid identification of C. difficile ribotypes is crucial but challenging.

Purpose of the Study:

• To develop a rapid MALDI-TOF MS workflow with machine learning for distinguishing epidemic C. difficile ribotypes (RT027 and RT181).
• To enable real-time identification of toxigenic C. difficile strains.

Main Methods:

• Analysis of MALDI-TOF MS spectra from 379 clinical C. difficile isolates.
• Identification of discriminant biomarker peaks using machine learning.
• Implementation of classifiers on the Clover MSDAS platform and the AutoCdiff web tool.

Main Results:

• Seven discriminant biomarker peaks were identified.
• Two peaks uniquely identified RT027; five peaks identified RT181.
• Classifiers achieved up to 100% balanced accuracy in ribotype assignment and demonstrated robustness in simulations.

Conclusions:

• MALDI-TOF MS coupled with machine learning provides rapid, high-precision C. difficile ribotype identification.
• The AutoCdiff tool offers a deployable solution for clinical labs to improve outbreak surveillance and control.