Application of MALDI-TOF MS-based peptidome profiling for the identification of Bacillus cereus, Staphylococcus aureus, and Escherichia coli in single and mixed inoculum

Area of Science:

• Microbiology
• Analytical Chemistry
• Bioinformatics

Background:

• Accurate detection of mixed bacterial species is crucial for food safety.
• Current methods for identifying mixed bacterial samples are limited.
• MALDI-TOF MS offers potential for bacterial identification but requires enhancement for complex samples.

Purpose of the Study:

• To develop an AI-driven approach for accurate identification of mixed bacterial species using MALDI-TOF MS peptidome profiles.
• To improve the accuracy and efficiency of bacterial strain detection in food safety applications.

Main Methods:

• Integration of Matrix-Assisted Laser Desorption/Ionization-Time of Flight Mass Spectrometry (MALDI-TOF MS) with Artificial Intelligence (AI).
• Utilized formic acid extraction and a residual network (ResNet) for improved prediction accuracy.
• Employed label-free proteomics via liquid chromatography-tandem mass spectrometry (LC-MS/MS) for peptidome profiling annotation.
• Applied a random forest-convolutional neural network (RF-CNN) for differential protein screening.

Main Results:

• Achieved 96.88% accuracy in identifying mixed bacterial samples containing Bacillus cereus, Staphylococcus aureus, and Escherichia coli.
• The combined approach of formic acid extraction, ResNet, and RF-CNN significantly enhanced strain identification in complex datasets.
• Label-free proteomics and differential protein screening provided further improvements in detection accuracy.

Conclusions:

• The developed AI-integrated MALDI-TOF MS strategy offers a robust method for accurate mixed bacterial species detection.
• This approach enhances the reliability of MALDI-TOF MS for food safety applications.
• Further research can expand this methodology to a wider range of bacterial species and food matrices.