Abstract Rapid and accurate identification of bacterial pathogens remains a major clinical challenge, particularly for distinguishing closely related species and resolving polymicrobial infections. Here, we present CRISPR-assisted Nanodroplet Differential Identification (CANDI), a high-throughput platform that combines species-agnostic target amplification with species-specific Cas12a detection in fluorescence-barcoded nanodroplets. Using conserved 16S/23S rRNA loci, we designed guide RNAs that enabled discrimination of 15 clinically relevant nontuberculous mycobacteria (NTM) species and subspecies. CANDI achieved 10 copies/reaction sensitivity, resolved minority strains in mixed infections, and distinguished M. abscessus subspecies using rationally engineered gRNAs. Applied to 300 patient-derived samples (L-J cultures, MGIT supernatants, and sputum), the platform delivered 95% accuracy within 4 hours. These results establish CANDI as a rapid, scalable, and clinically validated diagnostic technology for NTM and other challenging pathogens. This abstract is funded by: NIH
B Ning (Fri,) studied this question.