Ocular Microbiology and Immunology Group
Back to OMIG Main Page

< Previous | 2026 Agenda and Abstracts | Next >

2026 OMIG Abstract

Real-Time Adaptive Long-Read DNA Sequencing for Rapid Diagnosis of Microbial Keratitis

Russell N. Van Gelder, Marcus A. Toral, Kenji Nakamichi, Behrouz Rahimi, Marcus L. Turner,
and Miel Sundararajan

Karalis Johnson Retina Center, Department of Ophthalmology, University of Washington School of Medicine, Seattle, Washington

Purpose: To compare adaptive nanopore metagenomic DNA sequencing with standard microbiological culture for pathogen identification and time-to-result in infectious keratitis.

Methods: Prospective single-center diagnostic study of 47 corneal ulcer specimens (47 eyes, 46 subjects) with parallel clinical culture (plus Gram stain and PCR) and adaptive nanopore metagenomic sequencing on the Oxford Nanopore MinION. Adaptive sampling depleted human reads in real time; organisms were called at 1% genome breadth against the SMART database. Agreement and time-to-result (TTR) were compared, with TTR analyzed by Mann-Whitney U test.

Results: A pathogen was identified in 21/47 (45%) by culture and 24/47 (51%) by sequencing, with agreement in 36/47 (77%). Discordant results (11/47) were mostly low-biomass specimens positive by only one method, with one organism-level disagreement (diphtheroids vs. Finegoldia magna). Predominant organisms were P. aeruginosa, S. epidermidis, and S. pneumoniae; genome-normalized burden spanned ~4 orders of magnitude, with S. pneumoniae samples showing highest burden. Sequencing was significantly faster to result than culture (16.25 + 3.03 vs. 40.25 + 4.77 h; p = 0.0008), with adaptive sampling improving TTR ~1.5-fold.

Conclusions: Adaptive nanopore sequencing produced diagnostic yield comparable to culture while shortening organism identification by approximately one day. In addition to unbiased pathogen detection, the method quantified infectious burden across over four orders of magnitude. Genome-normalized burden may provide a clinically useful measure of result confidence and disease biology, particularly in high-burden pneumococcal keratitis, but requires prospective validation.