Skip to main content
. 2023 Sep 18;9(9):e20164. doi: 10.1016/j.heliyon.2023.e20164

Fig. 8.

Fig. 8

KRT81+ tumor cells phenotype can be used as a prognostic biomarker in patients with lung adenocarcinoma. (A) Identification of differentially expressed genes in two groups. (B) Random Forest algorithm to evaluate the effect of DEGs on survival. (C) The intersection of univariate COX result and the top ten importance in random forest identified robust prognosis-related genes. (D–E) Further screening of prognosis-related genes by LASSO algorithm. (F) Univariate Cox analysis of Risk Score and clinical index in the TCGA-LUAD cohort. (G) Multivariate Cox analysis of Risk Score and clinical index in the TCGA-LUAD cohort. (H) ROC curves of the predictive ability of the prediction model for the 1, 2, and 3-year prognosis in the training dataset (TCGA-LUAD). (I) ROC curves of the predictive ability of the prediction model for the 1, 2, and 3-year prognosis in the testing dataset (GSE30219 and GSE31210). (J) Constructing a nomogram for prognosis prediction in the TCGA-LUAD dataset. (K) ROC curves of risk score and clinical information on prognosis prediction. (L–N) Calibration curves of the nomogram at 1, 2, and 3 years.