A Machine Learning-Enabled SERS Sensor: Multiplex Detection of Lipopolysaccharides from Foodborne Pathogenic Bacteria

Area of Science:

• Food safety
• Biosensing
• Microbiology

Background:

• Foodborne pathogenic bacteria cause millions of illnesses and deaths globally each year.
• Lipopolysaccharides (LPS) are key biomarkers for Gram-negative pathogenic bacteria.
• Rapid and cost-effective pathogen detection is critical for public health.

Purpose of the Study:

• To develop a surface-enhanced Raman scattering (SERS)-based sensing platform for detecting and differentiating lipopolysaccharides (LPS) from pathogenic bacteria.
• To utilize linear polymer affinity agents and machine learning for enhanced bacterial detection.
• To assess the sensor's performance in complex food matrices.

Main Methods:

• Immobilization of linear polymer affinity agents on plasmonic substrates.
• Detection of LPS using surface-enhanced Raman scattering (SERS).
• Application of machine learning algorithms (PCA, SVM) for data analysis and bacterial discrimination.

Main Results:

• Successful characterization and differentiation of LPS from *Salmonella typhimurium*, *Escherichia coli* O111:B4, and *Escherichia coli* O26:B6.
• Demonstrated sensitive detection of LPS using the SERS platform.
• Validated the sensor's capability to detect LPS in apple juice, a complex food matrix.

Conclusions:

• A linear polymer-based SERS sensing platform integrated with machine learning is feasible for sensitive pathogen detection.
• This approach enables rapid and accurate differentiation of bacterial strains based on LPS biomarkers.
• The developed sensor shows significant potential for real-world food safety monitoring applications.