Genomic outbreak investigation of biosafety-level-3 pathogens using nanopore sequencing.

Background. Accurate outbreak analysis is essential for effective infectious disease control. While short-read Illumina sequencing is the current gold standard for genotyping pathogens, Oxford Nanopore Technologies (ONT) offers advantages such as portability and real-time sequencing. However, the accuracy of ONT for single nucleotide polymorphisms (SNP) detection and core genome multilocus sequence typing (cgMLST) remains poorly characterized, which is a critical issue when investigating outbreaks of highly dangerous biosafety-level-3 (BSL-3) agents.Aim. This study evaluates the potential of ONT sequencing for outbreak analysis of four BSL-3 bacterial species with low mutation rates: Bacillus (Ba.) anthracis, Brucella (Br.) melitensis, Brucella (Br.) suis and Francisella (F.) tularensis. Strains originating from epidemiologically defined outbreaks, with existing Illumina sequencing data, were selected for analysis. SNP calling was evaluated using three analytical strategies (PACU, clair3 and an assembly-based approach with snippy). The resulting genomic clusters were compared to epidemiologically defined outbreaks that had previously been reproduced using Illumina sequencing. Additionally, cgMLST was performed to evaluate genotyping resolution across the sequencing platforms.Results. Minor discrepancies in strain clustering were observed, particularly for F. tularensis, where homopolymeric regions contributed to false positives in read-based callers. The assembly-based snippy approach achieved the highest F1 score (0.96-0.99) across all species, followed closely by ab initio SNP callers PACU (F1 : 0.88-0.97) and clair3 (F1 : 0.83-0.98). Minor discrepancies in strain clustering were observed, particularly for F. tularensis, where homopolymeric regions contributed to false positives in read-based callers. CgMLST analysis showed high concordance between ONT and Illumina for Brucella spp., but greater variability for Ba. anthracis and F. tularensis.Conclusions. ONT sequencing, particularly when used with ONT assembly-based SNP calling, enables reliable outbreak analysis of highly pathogenic, low-diversity bacteria. While challenges remain for specific species and genomic features, ONT is a promising alternative for high-resolution bacterial genotyping in outbreak scenarios.