Reticular-Induced Energy Transfer Driven Renewable ECL System with Machine Learning for Glioma-Specific Dual-Biomarker Detection and Expression Correlation Mechanism

Rapid, accurate, and renewable electrochemiluminescence (ECL) bioassays are crucial for multiplexed biomarker detection. Integrated with efficient analytical model for processing sensing data, these tools enable precise differentiation of tumor stages and mechanistic insights into disease pathogenesis. In this study, we developed an ECL platform based on reticular-induced directed energy transfer by coordinating pyrene-derived covalent organic frameworks (pyr-COFs) with Pd²⁺ ions. Notably, the pyr-COFs functioned as energy transfer ligands rather than emitters, redirecting the ECL emission pathway from an oxidation-reduction process to a metal-to-ligand charge transfer (MLCT) route. This innovation achieved a 2.48-time increase in ECL efficiency compared to standard Ru(bpy)₃²⁺ system, enabling ultrasensitive detection. Building on this advancement, we constructed an automated and renewable ECL biosensor for sequential analysis of dopamine (DA) and miRNA-21. Also, this automated bioassay was applied to elucidate the correlation of DA depletion with miRNA-21 upregulation in glioma and the expression mechanism, achieving glioma staging. To enhance staging performance, a logistic regression-based machine learning algorithm was utilized, showing 100% accuracy in classifying healthy controls and low- and high-grade cases. This work provides instructive insights into development of next-generation renewable ECL biosensors and data algorithm models, paving the way for early glioma diagnostics, tumor staging, and pathogenesis research.