Gold Nanocluster-Based Sensor Array with PCA-Assisted Pattern Recognition for Discriminating Multiple Antibiotics in Different Complex Biological Samples

Principal component analysis (PCA)-assisted pattern recognition combined with sensor arrays shows great potential for direct antibiotic identification in complex biological samples, addressing critical needs in health monitoring and food safety. However, conventional approaches often require cumbersome surface modifications of multiple functional nanomaterials as nonspecific recognition units, increasing complexity and cost while hindering programmable array design. Here, we present a label-free sensor array based on gold nanoclusters (Au NCs) that enables the rapid discrimination of multiple antibiotics in various biological matrices. Notably, our system achieves antibiotic identification through the simple modulation of just four ligand molecules on Au NC surfaces, significantly simplifying programmable sensor design. Through comprehensive molecular docking simulations and fluorescence resonance energy transfer (FRET) experiments, we systematically elucidated the antibiotic response mechanism. The array demonstrates 100% accuracy in blind tests with human serum, urine, and milk samples. Most remarkably, PCA-assisted pattern recognition enables simple and rapid (3 min) discrimination between antibiotic-containing and antibiotic-free complex biological samples, highlighting its practical potential for life science and food safety applications.