Fig. 2From: Leveraging TCGA gene expression data to build predictive models for cancer drug responseAccuracy of random forest by number of clusters (using clara clustering algorithm). a Mean accuracy (200x cross-validation) was calculated using 1–204 clusters in order of importance from the 5-FU pan-cancer model. b Mean accuracy (200x cross-validation) was calculated using 1–192 clusters in order of importance from the GCB pan-cancer modelBack to article page