Fig. 1: Cladogram of 233 ColV-like plasmids (ColVLPs).

Unique ORFs from 233 ColVLPs were combined to form a hypothetical plasmid. Using sequence similarity searches against the hypothetical plasmid at an 80% nucleotide sequence identity and alignment length threshold, a binary sequence indicating open-reading frame presence/absence for each plasmid was generated. Binary sequences were subjected to hierarchical clustering using Manhattan distance and visualised as a cladogram. Plasmids were arranged into clusters (n > 1) based on the total within sum of squares method, with minor (n ≤ 5) and major clusters (n > 5). Host isolate source, host species, replicon type, predicted conjugative ability and the presence of colicins, colicin immunity genes (imm), virulence and resistance features are annotated. Feature prevalence is illustrated in a bar plot above the metadata. Potential conjugative ability was predicted based on the presence of all genes required for F plasmid conjugation⁸² determined using BLASTn at a 90% nucleotide sequence identity and alignment length threshold. Replicons were identified using ABRicate v0.8 against the PlasmidFinder database using a 100% query length threshold. Virulence genes were identified using a BLASTn search at a 90% nucleotide sequence identity and 95% alignment length threshold, with operons being considered present only if all genes were individually identified. Antimicrobial resistance genes were identified using ABRicate v0.8 against the ARG-ANNOT database⁸⁴ using a 100% query length threshold and grouped based on antibiotic class, with the following abbreviations: AGly aminoglycosides, Bla beta-lactamases, Fcyn fosfomycin, Flq fluoroquinolones, MLS macrolide-lincosamide-streptogramin, Phe phenicols, Rif rifampin, Sul sulphonamides, Tet tetracyclines and Tmt trimethoprim. Source data are provided as a Source data file in Supplementary Data 4.