Identification of Multiple Foodborne Pathogens Using Single-Atom Nanozyme Colorimetric Sensor Arrays and Machine Learning

Food-borne pathogens in food are a major hidden danger to human health. Rapid and accurate detection of food-borne pathogens is of great significance to people's food safety. However, current techniques are generally time-consuming and require large-scale equipment. Herein, a colorimetric sensor array combined with machine learning-assisted detection technology was developed, which can simultaneously and rapidly detect multiple foodborne pathogens. Firstly, Fe-N-C single-atom nanozymes (SAzymes) with high peroxidase-like activity were synthesized. The Fe-N-C SAzymes with strong signal amplification can specifically catalyze three kinds of chromogenic substrates, the colorless TMB, OPD and light green ABTS into blue oxidized TMBox, yellow oxidized OPDox, and green ABTSox, respectively. The different foodborne pathogens have different inhibitory effects on the active sites of Fe-C-N SAzyme, generating different colorimetric response signals according to three chromogenic substrates. The key aspect is that, through machine learning, colorimetric response signals were developed as unique fingerprints for various foodborne pathogens, enabling effective differentiation between them.The colorimetric sensor array based on SAzymes overcomes the limitations of traditional "lock−key mode" biosensors and achieves highly sensitive identification of multiple foodborne pathogens in the range from 105 to 108 CFU/ml. Furthermore, the constructed colorimetric sensor array assisted by machine learning shows good recognition performance in the detection of actual samples such as seawater, lake water and tap water, and can achieve simultaneous detection of five kinds of foodborne pathogens, indicating great potential for the rapid, precise and high-throughput identification of foodborne pathogens in the field of food safety.