A novel intelligent sensing strategy: Integration of metal-doped carbon dots nanozymes and machine learning for rapid screening of biothiols in disease

Abstract

Biothiols play a crucial role in signal transduction and cellular metabolism, and their accurate detection is essential for biomarker monitoring and disease diagnosis. However, due to the similar chemical structures of different biothiols, traditional detection methods often suffer from false positive signals, mutual interference, expensive instrumentation, and complex operations. Array sensing technology provides an ideal solution as an efficient multi-channel parallel detection method that can analyze multiple targets simultaneously. A nanozyme-based colorimetric sensor array is proposed with the aim of providing a high-throughput, high-sensitivity, high-selectivity, low-cost, and convenient solution for the detection of biothiols. Metal ion-doped carbon dots (M-CDs) were employed as sensing units to mimic peroxidase activity, catalyzing the oxidation reaction of 3,3′,5,5′-tetramethylbenzidine (TMB). Different biothiols exhibited varying degrees of inhibition on the catalytic activity of M-CDs, resulting in unique fingerprint patterns. The method was integrated with various machine learning algorithms, including linear discriminant analysis (LDA), hierarchical clustering analysis (HCA), artificial neural networks (ANN), K-nearest neighbors (KNN), decision trees (DT), and support vector machines (SVM), accurately distinguished eight biothiols with a detection limit of 40 nM. This approach not only overcomes the limitations of traditional techniques but also enables efficient detection in complex biological samples (such as serum, urine, cells, and bacteria), providing a sensitive and straightforward technological platform for biothiol monitoring in medical diagnostics.