Fig. 3: Varied RplD macrolide binding site mutations are associated with increased azithromycin MICs.

Mean (a) and distribution (b) of azithromycin MICs for RplD macrolide binding site variants. Previously reported mutations are cited with the first reporting publications. Violin and swarm plots and statistical analyses were limited to isolates with MICs < 8 to exclude isolates with 23S rRNA mutations. Rare RplD mutations (denoted as “Other (n = 31)”) were grouped for visualization and statistical analysis, and thus were assumed to all have the same effect. Quartiles within violin plots are depicted using dotted lines. Statistical significance between RplD variants and RplD wildtype MIC distributions was assessed by two-sided Mann–Whitney U Test: *p < 0.05, **p < 0.01, and ***p < 0.001. Exact p-values from left to right were 3.09 × 10−7 (WT vs. G70D), 3.91 × 10−10 (WT vs. Other), 5.74 × 10−5 (G70D vs. Other).