Supplemental dataset 1. Raw metadata of downloaded genomes from PATRIC containing all the data associated with each genome.
Supplemental dataset 2. Species by species tabulation and analyses of the resistomes, specimen sources, country of isolation, MLST, Biosample accession number, and strain name of all the genomes according to their order on the phylogeny trees.
Supplemental dataset 3. Colour‐coded species by species tabulation of the resistomes, specimen sources, country of isolation, MLST, Biosample accession number, and strain name of all the genomes according to their order on the phylogeny trees.
Supplemental dataset 4 Plasmid replicons and their associated resistomes and bacterial hosts. Selected genomes bearing carbapenemases, mcr, fluoroquinolones, aminoglycosides and other clinically important antibiotic resistance genes, were run through PlasmidFinder, pMLST, and BLASTn to identify their genetic environment. Most antibiotic resistance genes were found on plasmids, making them potentially mobile.
Supplementary Figures: Figure S1A‐B Evolutionary epidemiology and resistome of global Citrobacter freundii, brakii, portucalensis, and amalonaticus isolates. S1A shows Citrobacter sp., particularly freundii, portucalensis and brakii clustering into clades A, B1, B2 and B3 whilst S1B shows C. amalonaticus strains clustering into clades A (red highlight), B1 (green highlight), B2 (blue highlight) and C (mauve highlight); clade B2 had very rich resistome repertoire and were all from France, but the other clades had very few resistance genes. Strains from humans (blue labels), animals (red labels), plants (purple/mauve labels) and the environment (green labels) were found in the same clade/cluster. Included in S1A are Pseudomonas, Klebsiella, and Escherichia coli species that were originally classified as C. freundii but later reclassified into their actual species using ANI; their clustering away from the Citrobacter species confirms the ANI results that they were initially misclassified. BlaSED and oqxAB were almost conserved in these genomes. Branches with bootstrap support values of ≥50 were defined as belonging to the same clade. The branch lengths also show the evolutionary distance between the isolates. Blue and red arrows show the direction of evolution as well as local and international dissemination of strains of the same clone/clade through different hosts.
Fig. S1C. Evolutionary epidemiology and resistome of global Citrobacter koseri isolates. C. koseri strains clustered into clades A (grey highlight), B1 (light blue highlight), B2 (orange highlight) and B3 (mauve highlight). Strains from humans (blue labels) and animals (red labels) were found in the same clade/cluster. BlaCKO and blaMAL were almost conserved in these genomes. Included in S1C are Serratia marcescens, Klebsiella, Enterobacter and Escherichia coli species that were originally classified as C. koseri but later reclassified into their actual species using ANI; their clustering away from Citrobacter koseri confirms the ANI results that they were initially misclassified. Branches with bootstrap support values of ≥50 were defined as belonging to the same clade. The branch lengths also show the evolutionary distance between the isolates. Blue and red arrows show the direction of evolution as well as local and international dissemination of strains of the same clone/clade through different hosts.
Fig. S1D. Evolutionary epidemiology and resistome of global Citrobacter spp, isolates, A and B. This tree shows information for additional Citrobacter freundii and Citrobacter sp. that were not featured figures 1, and S1A–C above. Included in S1D are Serratia marcescens, Klebsiella, Enterobacter and Escherichia coli species that were originally classified as C. freundii, but later reclassified into their actual species using ANI; their clustering away from C. freundii confirms the ANI results that they were initially misclassified. C. freundii clustered into four main clades (A, B1, B2 and B3), highlighted with distinct colours. Clade B3 had the most resistome abundance and diversity. Strains from humans (blue labels), animals (red labels), plants (purple/mauve labels) and the environment (green labels) were found in the same clade/cluster. BlaCMY was conserved in these genomes. Branches with bootstrap support values of ≥50 were defined as belonging to the same clade. The branch lengths also show the evolutionary distance between the isolates. Blue and red arrows show the direction of evolution as well as local and international dissemination of strains of the same clone/clade through different hosts.
Figs S2A‐B. Evolutionary epidemiology and resistome of global Enterobacter hormaechei subsp. Hormaechei, Xiangfangensis, Oharae, and Steigerwaltii, isolates. S2A is strictly E. hormaechei subsp. Hormaechei and is an addition to Figure 4 whilst S2B is an addition to Figures 3‐4 above as additional genomes of E. hormaechei subsp. Xiangfangensis, Oharae, and Steigerwaltii; these could not be added to Figures 3‐4 and are shown here in Fig. S2B. The E. hormaechei isolates in S2A clustered into three main clades A, B and C (with distinct highlights) that contained strains distributed globally from humans (blue labels), and animals (red labels), plants (purple/mauve labels) and the environment (green labels). Clades B and C contained diverse and rich resistome repertoire. blaACT was conserved in these genomes. S2B contains E. hormaechei subsp.Xiangfangensis, Oharae, and Steigerwaltii isolates clustering into 6 branches (I‐VI); genomes of the same subsp. clustered closely together. Branches with bootstrap support values of ≥50 were defined as belonging to the same clade. The branch lengths also show the evolutionary distance between the isolates. Blue and red arrows show the direction of evolution as well as local and international dissemination of strains of the same clone/clade through different hosts.
Fig. S3A‐B. Evolutionary epidemiology and resistome of global Klebsiella variicola isolates. S3A‐B are additional trees to Figure 5 and show additional K. variicola genomes that could not be added to Figure 5; in all, Figures 5 and S3A‐B show 600 K. variicola genomes. S3A and S3B trees are composed of different K. variicola genomes, which is reflected in the differences in the resistomes and tree topologies. Included in S3A and S3B are K. pneumoniae and K. pneumoniae and quasipneumoniae species respectively, that were originally classified as K. variicola, but later reclassified into their actual species using ANI; their clustering away from K. variicola confirms the ANI results that they were initially misclassified. The K. variicola strains clustered into eight (S3A) and seven (S3B) clades I‐VIII and I‐VII respectively, which were highlighted with distinct colours and were isolated from countries around the globe. The clades contained strains distributed globally from humans (blue labels), animals (red labels), plants (purple/mauve labels) and the environment (green labels). Besides a few strains in clade B2, the other strains contained very few resistance genes. blaLEN was conserved in these genomes. Branches with bootstrap support values of ≥50 were defined as belonging to the same clade. The branch lengths also show the evolutionary distance between the isolates. Blue and red arrows show the direction of evolution as well as local and international dissemination of strains of the same clone/clade through different hosts.
Fig. S4. Evolutionary epidemiology and resistome of global Proteus mirabilis isolates. The P. mirabilis isolates clustered into 10 clades, A‐A3, B1‐B3, and C1‐C3 (shown with different highlights), which contained diverse and abundant resistomes with conserved catA and tet genes. The clades contained strains distributed globally from humans (blue labels), animals (red labels), plants (purple/mauve labels) and the environment (green labels). Branches with bootstrap support values of ≥50 were defined as belonging to the same clade. The branch lengths also show the evolutionary distance between the isolates. Blue and red arrows show the direction of evolution as well as local and international dissemination of strains of the same clone/clade through different hosts.
Fig. S5 (A‐C). Count of antibiotic resistance genes (ARGs) in Citrobacter freundii (A), and Citrobacter species (B and C). The sum of each unique ARG and its variants are computed and shown as a bar chart to depict the most abundant ARGs.
Fig. S6 (A‐B). Count of antibiotic resistance genes (ARGs) in Citrobacter amalonaticus (A), and Citrobacter koseri (B). The sum of each unique ARG and its variants are computed and shown as a bar chart to depict the most abundant ARGs.
Fig. S7 (A‐B). Count of antibiotic resistance genes (ARGs) in Enterobacter steigerwaltii and oharae (A), and Enterobacter xiangfangensis (B). The sum of each unique ARG and its variants are computed and shown as a bar chart to depict the most abundant ARGs.
Fig. S8 (A‐C). Count of antibiotic resistance genes (ARGs) in Enterobacter hormaechei (A, B and C). The sum of each unique ARG and its variants are computed and shown as a bar chart to depict the most abundant ARGs.
Fig. S9 (A‐C). Count of antibiotic resistance genes (ARGs) in Klebsiella variicola (A, B and C). The sum of each unique ARG and its variants are computed and shown as a bar chart to depict the most abundant ARGs.
Fig. S10 (A‐C). Count of antibiotic resistance genes (ARGs) in Morganella morganii (A), Proteus mirabilis (B) and Providencia species (C). The sum of each unique ARG and its variants are computed and shown as a bar chart to depict the most abundant ARGs.
Table S4. Statistical analyses of resistome diversity, abundance, and relative richness
