Ultrasensitive Detection via Machine Learning-Optimized Bacterial-Imprinted Photoelectrochemical Biosensor with Active/Passive Dual-Mode Validation.

Currently, the existing detection platforms face persistent challenges in achieving reliable bacterial identification within complex matrices, particularly in food and environmental specimens, where matrix interference effects substantially compromise analytical sensitivity. Herein, we utilize a combination of and 4-ethynacetophenone (4-EAP, functional monomer) to create a novel "Bidirectional" BIPs-PEC biosensor and take advantage of bacterial imprinting technology (BIT) and photoelectrochemical (PEC) to detect with precise identification and high sensitivity detection. Under positive bias voltage (active mode), as the cavities in the BIPs were occupied by , it drives the electrostatic interaction between 4-EAP and the negatively charged cell membrane of and hydrophobic interactions, facilitating the effective transfer of multitudinous photo-generated electrons to Conversely, negative bias application triggers passive mode operation, where the negatively charged electrode surface generates electrostatic repulsion against cells while suppressing alkyne group reactivity, hindering electron transfer. The results obtained using the two modes are verified against each other, effectively minimizing errors and interference from the background signal. It demonstrated significant selectivity for compared to other bacteria and retained superior efficacy in intricate food matrices, identifying at concentrations as low as 10 CFU/mL. The preparation and detection of sensors were analyzed and predicted by using molecular docking and machine learning. This work significantly mitigates the impact of interference factors in the conventional mode, providing numerous benefits, including convenience and speed, efficiency, and accuracy, and holds a highly promising method for microorganism detection in food and environmental domains.