Plasmid typing challenges in the era of genomic antimicrobial resistance surveillance.

Abstract

Plasmids are central drivers of antimicrobial resistance (AMR) dissemination, yet their modular, recombinogenic genomes defy concepts of relatedness derived from bacterial chromosomes. Long-read sequencing now routinely yields complete plasmid sequences, exposing a key gap: we can reconstruct plasmids but lack agreed principles for comparing them. In response, diverse clustering tools have emerged that emphasise different signals, such as marker genes, whole-sequence similarity, gene content, backbone structure, or explicit rearrangement events. Each encodes a distinct notion of plasmid distance. This review organises these tools into conceptual families, highlights how their assumptions shape surveillance outputs, and proposes hierarchical, multitool strategies and clearer models of plasmid evolution as the basis for scalable and interpretable plasmid-based AMR surveillance.