Machine learning-empowered nanozyme-based aptasensor arrays for accurate discrimination and sensitive quantification...

Kynurenine pathway (KP) metabolites are closely associated with physiological homeostasis and pathological progression, serving as pivotal disease biomarkers. However, efficient simultaneous discrimination and quantification of these structurally analogous metabolites remain a persistent challenge in analytical biology. Herein, we developed an accurate detection platform for these metabolites by integrating aptamers (Apt), iron-centered porphyrin aromatic framework-based nanozymes (PAF(Fe)) and machine learning algorithms. Specifically, PAF(Fe) was individually conjugated to five distinct aptamers targeting KP metabolites, yielding five PAF(Fe)-Apt conjugates that were assembled into an aptasensor array. Upon binding to their cognate aptamers on PAF(Fe)-Apt, target KP metabolites suppressed the peroxidase-like activity of PAF(Fe), thereby inducing specific signal attenuation. The array demonstrated a broad linear range of 0.01-2.0 μg/mL and low limits of detection for all five KP metabolites. Moreover, linear discriminant analysis, hierarchical cluster analysis and random forest algorithms achieved 100% discriminative accuracy, while a dual-output artificial neural network model facilitated the simultaneous qualitative and quantitative detection of these metabolites. Additionally, the platform not only enabled 100% precision classification of multivariate KP metabolite mixtures but also exhibited no cross-reactivity toward structurally analogous interferents and strong anti-interference ability against metal ions and biomolecules. Moreover, the platform delivered reliable analytical performance in serum sample assays with recoveries ranging from 98.39% to 106.65%, and could be integrated with a user-friendly web interface for real-time result output. Overall, this work established a robust tool for comprehensive KP metabolite profiling, holding substantial potential in the early diagnosis of diseases and related pathological research.