Skip to main content
. 2018 Sep 12;21(3):277–287. doi: 10.4048/jbc.2018.21.e39

Figure 3. The prediction performance of the tamoxifen efficacy-related long noncoding RNA (lncRNA) signature (TLS) in the discover dataset after optimization. The TLS was optimized with best cutoff value. After that, the performance of TLS and expression profile of lncRNAs in TLS was analyzed in the discover dataset. (A) In the discover dataset, receiver operating characteristic (ROC) curve for the performance of TLS in distant metastasis-free survival (DMFS) was plotted with the corresponding area under the ROC curve (AUC) and the best cutoff for TLS score was determined. (B) The expression profile of the six lncRNAs in TLS, the risk score of TLS and patients' DMFS were integrated and then evaluated in the discover cohort. (C) Patients classified by TLS with optimized cutoff value were evaluated in Kaplan-Meier analysis in the discover dataset.

Figure 3

HR=hazard ratio.