A group from Department of Chemistry, Pavillon Alexandre-Vachon, 1045, avenue de la Médecine, Université Laval, Quebec, Canada has reported Lipopolysaccharide (LPS) detection and its bacteria identification with using a panel of several lectins and by applying a machine learning method on it.
https://pmc.ncbi.nlm.nih.gov/articles/PMC12019741/
The sensing and classification of LPS, which are pivotal constituents of Gram-negative bacteria, are fundamentally important in fields such as healthcare, environmental monitoring, and food safety.
This study presents a new approach utilizing a panel of lectins (from 2 to 7 kinds) immobilized on surface plasmon resonance (SPR) sensors. Each glycan binding profile, which is unique to the bacteria, was used to identify the species of bacteria combining with a machine learning method with high accuracy. Used machine learning methods were Random Forest (RF), k-Nearest Neighbors (kNN), and Support Vector Machine (SVM).
It seems that this kind of detection method using multi-probes combined with a machine learning method is a recent trend in sensing applications.