A novel sensor array combined with machine learning for the effective differentiation of multiple classes of antibiotics based on copper-based nanozymes with laccase-like activity.

A nanozyme sensor array was developed combining with machine learning algorithms for the effective identification and differentiation of multiple antibiotic classes and antibiotic individuals. The four channel sensor array was composed by four copper-based nanozymes with laccase-like activity. The significant structural differences between different classes of antibiotics (such as Aminoglycosides, Tetracyclines, β-lactams, Nitroimidazole, Sulfonamides, Quinolones) resulted in the different effect on the nanozymatic reactions through interact with copper、substrate and so on. In addition, the differences between antibiotics in the same class also made certain effect on the nanozymatic reactions. Based on the above affect and differences, we combined the machine learning algorithms to successfully realize the identification of six major classes and sixteen individuals of antibiotics simultaneously for the first time. The proposed sensing strategy realized the facile identification of major categories of antibiotics, which has good application prospects in the rapid antibiotic identification in environmental water bodies.