Skip to main content


Fig. 2 | Genome Medicine

Fig. 2

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

Fig. 2

Methodological pipeline to identify drug modules. (A) The process of obtaining modules starts with the calculation of z-transformed Pearson’s correlation (zcor) between gene expression and drug sensitivity data for each drug-gene pair. Correlations (zcor) beyond ± 3.2 are considered to be significant. (B) We then run a gene-set enrichment analysis (GSEA) for each drug in order to identify genes that participate in enriched Reactome pathways. (C) GSEA-filtered genes are submitted to HotNet2 on the STRING interactome in order to identify drug modules. (D) Finally, modules are expanded (when possible) using the DIAMOnD algorithm to recall the few correlated genes that might have been excluded in step C as a result of the limited coverage of the Reactome database. This final step has a minor impact on the composition of the module

Back to article page