Orthogonal Biochemical Sensing for Concentration-Independent Bacterial Fingerprinting

Array-based biosensors hold substantial promise for rapid bacterial identification. However, conventional approaches face two key limitations: their reliance on nonspecific interactions with bacterial surfaces hinders biochemical interpretation, and their detection of bulk populations results in concentration-dependent responses that obscure distinctive bacterial fingerprints. Here, we present a novel, concentration-independent bacteria identification (CIBI) sensor array strategy that profiles the intrinsic biochemical characteristics of bacteria at the single-cell level to generate distinct fingerprints. The array consists of three sensing modules: two unnatural d-amino acid probes targeting different enzymatic pathways in peptidoglycan synthesis, and a phenylboronic acid probe recognizing surface polysaccharides. By measuring per-cell fluorescence from ∼10,000 individually interrogated bacteria rather than population responses, the system achieves concentration-independent profiling. Combined with machine learning, the CIBI strategy accurately identifies nine bacterial strains (10⁵ CFU/mL) with 92.2% accuracy within 100 min. Remarkably, it also predicts the identity of previously unseen strains. In clinical applications, the array identified pathogen-spiked urinary tract infection samples with 95.2% accuracy, improving to 97.6% using a random forest algorithm. Overall, this strategy offers a robust platform for rapid and reliable clinical diagnostics.