Unveiling Diversity: Classification of Klebsiella Pneumoniae Plasmids from Long-read Assemblies

Plasmids, integral to bacterial evolution, pose challenges in their genome classification due to incomplete assembly data. While next-generation sequencing has improved plasmid classification, challenges persist in accurately assembling complete plasmid genomes. This study presents a novel plasmid classification methodology based on complete genome similarity, utilizing three metrics: nucleotide composition, gene occurrence, and structural dissimilarity. Tested on a local Klebsiella pneumoniae population, the method outperforms pMLST and PlasmidFinder, distinguishing plasmids even in fusion cases. Applied across diverse bacterial populations, this reference-free approach proves adaptable, offering a valuable tool for monitoring plasmid mobility and diversity. Third-generation sequencing advancements provide a comprehensive understanding of plasmid dynamics, which is essential for addressing antibiotic resistance and bacterial pathogenicity.