Fig. 9
Hierarchical clustering and correlation analysis of antimicrobial resistance determinants. This comprehensive heatmap presents a correlation matrix with hierarchical clustering of antimicrobial resistance genes and corresponding phenotypic resistance profiles. Both axes display the same set of determinants, organized through unsupervised hierarchical clustering based on Pearson correlation coefficients. The resulting dendrograms reveal phylogenetic relationships and functional associations among resistance elements. Genes and antimicrobial agents that exhibited complete absence or complete presence across all isolates were excluded from the analysis, as they would yield undefined or meaningless correlation coefficients and not contribute to the clustering pattern interpretation. Distinct clustering patterns emerged, with solid-line enclosures highlighting significantly correlated gene groups that suggest potential genetic linkage or co-regulation mechanisms. The prominent cluster in the upper left quadrant groups β-lactam resistance elements with strong intercorrelations, while the tet(M)-sul2 cluster (r = 0.54) indicates possible plasmid-mediated co-transfer. Weakly correlated elements appear in separate clusters, reflecting independent genetic origins. Color intensity represents correlation strength ranging from + 1 (deep red) to −1 (deep blue), with white indicating no correlation. The symmetric matrix structure enables simultaneous visualization of both gene-gene and gene-antibiotic relationships within a unified analytical framework, providing insights into the complex network of antimicrobial resistance determinants.
