Fig. 6: The R-MIC and its global dynamics for pathogen-antibiotic (PA) pairs are different from mean MIC dynamics shown in Fig. 5.
From: Seeking patterns of antibiotic resistance in ATLAS, an open, raw MIC database with patient metadata
A Example dynamics of the S-MIC (blue dash) and R-MIC (red dash) are shown for 4 PA pairs: these have 1, 2 or 3 clusters in their MIC distributions and linear regressions predict the changes of the R-MIC (red) and S-MIC (blue). B Predicted global R-MIC changes per year for each PA pair (\({\log }_{2}\) MIC change per year) are shown as a clustergram (the analogous plot with European data is similar, Supplementary Fig. 15B). C Changes in R-MIC (y-axis) and mean MIC (x-axis) are not well correlated (one dot per PA pair, ρ = 0.33, red dash shows `y=x' to highlight equal changes, crosshairs indicate s.d. from synthetic ATLAS replication, n = 50) because many PA pairs have R-MIC decreases and mean MIC increases and vice versa (c.f. Figs. 5B and 6B or Supplementary Figs. 15A and B). D These 4 PA pairs illustrate 4 MIC dynamics whereby R clusters have either sub- or super-breakpoint MICs which either increase or decrease (making 4 possible cases).
