Fig. 4.

LncRNA-based EN-prediction models predict drug response in patient tumors. a Imatinib (left), bleomycin (middle), and gefitinib (right) were predicted to be sensitive in FDA approved or yet-to-be approved cancer types. b Different overall survival rates between patients with different predicted responses to each of the 49 FDA approved cancer drugs in individual cancer type. Blue and red represented hazard ratio by the univariate cox regression model. The asterisks indicate the drugs that are approved by FDA for corresponding cancer types. c Forest plot of multivariate Cox regression analysis of “Drug resistance”, “Stage” and “Age at diagnosis” on patient survival in four cancer types. d The Kaplan–Meier curves of overall survival for OV, STAD, and UCEC patients who received corresponding treatments. The patients were segregated by median predicted IC50s