Qualitative and quantitative analysis of volatile metabolites of foodborne pathogens using colorimetric-bionic sensor coupled robust models

In a quest to safeguard humans, the detection of pathogenic bacteria is beneficial. Bacterial volatile metabolites are vital component of bacteria, that helps in communicating with the surrounding of various biological bacteria pheromone. Herein, the potential of colorimetric-bionic sensors (CBs) integrated with headspace solid-phase micro extraction-gas chromatography-mass spectrometry (HS-SPME-GC/MS) and multivariate calibrations for sensing analysis of volatile metabolites of E. coli has been explored. The CBs were fabricated based on chemoresponsive dyes and the colour characteristic fingerprints were collected and statistically analyzed at different incubation times using PCA, LDA, and HCA models. HCA model was positive in identifying the samples from six incubation times. Comparatively, based on qualitative identification of volatile metabolites in E. coli, the online sequential extreme learning machine (OSELM) achieved promising performance with a 96.67% discrimination rate. The study further revealed the correlations between volatile metabolites of E. coli and total bacterial count by robust modelling with a better performance by uninformative variable elimination-ELM (UVE-ELM) (Rp = 0.9405). Finally, a total of 35 metabolites were identified with a sensitive response to CBs according to the metabolite profiling. This affirmed the potential of the platform for flexible and versatile bacterial monitoring and detection.