Fig. 4

Quantitative scores of differential cellular response to perturbations. a An example of the predicted drug efficacy score. b (upper panel) Distribution of the predicted drug efficacy (D score). Each box plot shows the distribution of drug efficacy scores of 45 distinct networks. The bold lines and the boxes represent the median and the interquartile range (25th and 75th percentiles) and the whiskers extended to 1.5 times the interquartile range. If D score > 0.5, we consider the corresponding drug is effective. Dot color denotes major cellular phenotype of each differentially wired network (DWN) in response to perturbation (red: P, green: A, blue: D). (lower panel) Distribution of the effective drug targets (i.e., high drug efficacy, when death is the major cellular phenotype). Effective drug targets are divided into all altered gene target, combination of altered gene target, and wild-type gene target and all wild-type gene target. c An example of the predicted drug synergy score. Y-axis is the ratio of corresponding DWNs to total DWNs for the effective inhibitory treatment. d (upper panel) Distribution of the predicted drug synergy score (S score). S scores are converted to z-scores to facilitate comparison and definition of the synergistic (upper line)/antagonistic (bottom line) thresholds. Box plots show the median (the bold lines), the interquartile range (the boxes), and the lowest and highest scores within 1.5 times the interquartile range (the whiskers). Dot color denotes major cellular phenotype after perturbation (red: P, green: A, blue: D). (lower panel) Distribution of the selected synergistic/antagonistic drug combinations, consisting of altered gene target and wild-type gene target. Y-axis is the ratio of corresponding DWNs to total DWNs for the effective inhibitory treatment. e Selected synergistic or antagonistic drug pairs that comprise wild-type gene targets based on the drug synergy score. f Prediction of effective network-specific drug target(s) that comprise wild-type gene targets