Skip to main content
Fig. 1 | Genome Medicine

Fig. 1

From: Encircling the regions of the pharmacogenomic landscape that determine drug response

Fig. 1

Analysis of drug-gene correlations. a Observed drug-gene correlation distribution (purple) and randomized drug-gene correlation distribution (blue) (random permutation of expression values). Vertical lines denote the percentiles 5 and 95 of each distribution. b The left panel shows the “number of correlated genes per drug”, while the right panel shows the “number of correlated drugs per gene”. In the left panel, one can read, for example, that there are about 25 drugs (y-axis) with at least 1250 correlated genes (x-axis). Likewise, in the right panel, one can read that about 4000 genes (y-axis) are correlated to at least 10 drugs (x-axis). c Number of positively (red) and negatively (blue) correlated genes across drug classes. d Positively correlated targets (see the “Methods” section for details on the z-score normalization procedure of this correlation measure). Each dot represents one drug-target correlation. A full account of drug-target annotations is provided in Additional file 8. The red boxplot shows the background (random) distribution. e Drug-gene correlations (zcor) between afatinib/gefitinib and the epidermal growth factor receptor (EGFR) across tissues. In the upper plots, we show the drug sensitivity (1-AUC) across tissues. In the middle plots, we show basal gene expression of EGFR across tissues. Bottom plots show the Afatinib/Gefitinib-EGFR correlation. The rightmost values refer to the correlation when all tissues are considered (Global). Size of the bubbles is proportional to the number of CCLs in each tissue

Back to article page