FIGURE 7.

Analysis of prognostic assessment of BET genes in TCGA dataset (A–D), and external validation of BRD4 gene signature in ICGC training set (E–H). Patients were divided into low-risk and high-risk groups. (A) LASSO regression analysis was performed on BET genes to calculate the correlation coefficients. Coefficients of selected features are shown by lambda parameter; Partial likelihood deviance versus log(λ)was drawn using LASSO Cox regression model. (B) The LASSO algorithm was used to generate risk scores for the training cohort from TCGA. Relationship of BET proteins expression with risk score, survival time, and survival status were shown in the training cohort. (C) Distribution of KM survival curves for differential expression of BET proteins in the training cohort. (D) ROC curve and AUC of BET proteins signature classification. (E) Kaplan-Meier survival analysis of patients in different risk groups from TCGA dataset. (F) Kaplan-Meier survival analysis of patients in different risk groups from ICGC training set. (G) The ROC curves for risk score in TCGA dataset. (H) The ROC curves for risk score in ICGC training set. The higher values of AUC corresponding to higher predictive power.