Discriminant analysis of a mixture of pathogenic bacteria into different types and proportions by surface-enhanced Raman scattering spectroscopy combined with chemometric methods

Accurate identification and discrimination of bacteria is crucial for ensuring food safety and reducing pathogenic infections. This study presents a novel approach that combines surface-enhanced Raman scattering spectroscopy (SERS) with chemometric methods for discriminant analysis of a mixture of pathogenic bacteria into different types and proportions. Au@Ag@SiO2 composite nanomaterials were employed as the SERS substrate to collect Raman spectra of multiple pathogenic bacteria. Partial least squares discriminant analysis (PLS-DA) and orthogonal partial least squares discriminant analysis (OPLS-DA) methods were combined with standard normal variate (SNV) to discriminate the different species of mixed bacteria and the multiple proportion mixed bacterial samples, respectively. The results showed that SNV-PLS-DA had good classification performance in the discriminant analysis of different species of mixed bacteria, with an accuracy of 92% for the external test set. Furthermore, both SNV-PLS-DA and SNV-OPLS-DA models exhibited excellent classification performance in the discrimination of multiple pathogenic bacteria at different mixing proportions, achieving 100% accuracy in the external test set, but except for mixed samples of Escherichia coli and Salmonella typhimurium. Our method demonstrates the accurate capability of the SERS platform combined with chemometric methods in the discriminant analysis of multiple pathogenic bacteria at different species and mixing proportions, which provides novel insights for the synchronous analysis of multiple pathogenic bacteria.