Analytical validation of a highly accurate and reliable next-generation sequencing-based urine assay.

Urinary tract infections (UTIs) are diagnosed based on symptoms and confirmed by urine culture, despite its limitations in sensitivity. False-negative cultures can lead to inappropriate antimicrobial use or urosepsis in high-risk patients. Next-generation sequencing (NGS)-based metagenomics offers a comprehensive and precise alternative but is rarely applied clinically. We developed and validated BIOTIA-ID, a clinical-grade NGS-based diagnostic pipeline for pathogen detection in urine. Remnant clinical and spiked urine samples underwent extraction, metagenomic library preparation, and Illumina NextSeq 550 sequencing. We trained and applied a bioinformatic pipeline that uses machine learning to identify pathogens and resistance markers. BIOTIA-DX was intentionally designed and trained to increase stringency and reduce false positive detection of urogenital commensals or opportunistic microbes present at colonization levels. Internal controls ensured standardized, high-stringency results. The assay was validated on 1,470 urine specimens evaluating over 14.5k analytes. The clinical validation achieved a 97.2% sensitivity and 99.6% specificity with a limit of detection (LoD) of <15,000 CFU/mL for most bacterial species and <5,000 CFU/mL for fungal species. Discordant results were reconciled by target-specific qPCR or 16S Sanger sequencing, and 87% of the NGS results were concordant with the comparator. A subset of 332 clinical specimens was tested and validated for antimicrobial resistance (AMR).