Skip to main content
. 2019 Jul 16;42(8):579–588. doi: 10.14348/molcells.2019.0065

Fig. 3. Retrieval of known drugs for matched disease expression datasets using CMap and NGSEA.

Fig. 3

(A) Overview of how NGSEA retrieves drugs for the matched disease of the given gene expression data. Drug A targets genes with high network-based scores, whereas Drug B targets genes with a wide range of network-based scores; therefore, Drug A but not Drug B will be highly ranked. If Drug A but not Drug B is a known drug for the given disease, the prediction performance of NGSEA, as measured by the AUROC will be high. (B) Comparison of the AUROC between CMap and NGSEA for the ability to retrieve known drugs for the matched disease of each gene expression data set.