A Recurrence-Specific Gene-Based Prognosis Prediction Model for Lung Adenocarcinoma through Machine Learning Algorithm

Figure 3

Efficiency of RFS prediction model. (a) The Kaplan-Meier (K-M) curve confirmed that the signature could significantly distinguish low- and high-risk groups in the training cohort. (b) The K-M curve confirmed that the signature could significantly distinguish low- and high-risk groups in the internal validation cohort. (c) The K-M curve confirmed that the signature could significantly distinguish low- and high-risk groups in the external validation cohort (GSE68465). (d–g) The K-M curve confirmed that the prediction model could distinguish low- and high-risk groups in the pathological subgroups (d, e) and smoking history subgroups (f, g). (h) Forest plot showed results of multivariate cox analysis. (i) Receiver operating characteristic curve showed the prediction model obtained good predictive effect compared to other clinicopathological features.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)