Figure 4: Analysis of cross-resistance and trade-offs in resistance to different antibiotics using an antibiotic resistance network.

(A) A network of 339 antibiotic-resistant genes based on PseudomonasNet is shown. Blue nodes represent genes in which antibiotic resistance is decreased by knockout (i.e., positive regulation of antibiotic resistance) and red nodes represent genes in which antibiotic resistance in increased by knockout (i.e., negative regulation of antibiotic resistance). Yellow nodes represent genes that show a trade-off in resistance to different antibiotics. The node size is proportional to the number of antibiotics whose resistance are changed by perturbation of the gene. (B) Adherence of the genes with each direction of mutational effect to the same directional group of genes in PseudomonasNet. Box and error bars represent distribution of mean adherence score and standard deviation, respectively, for each group of genes. (C) The normalised connection scores to two groups of genes for different directions of mutational effect are represented as bars projecting in opposite directions: right for those that decrease antibiotic resistance by knockout and left for those that increase antibiotic resistance by knockout. If the given gene interacts with other genes with the same direction of mutational effect for each drug, then the blue bars are expected to project to the right and the red bars are expected to project to the left. The asterisk symbol (*) represents genes that exhibit adherence to the same direction of mutational effect. Genes that are underlined were shown to switch their interactions between directions of mutational effect in different antibiotic treatments. (D) The distribution of the number of genes that switch their interactions between directions of mutational effect under different antibiotic conditions for 1,000 randomised networks is shown.