Identification of the Characteristic Genes and their Roles in Lung Adenocarcinoma Lymph Node Metastasis through Machine Learning Algorithm

Figure 3

Prognosis model construction. Notes: (a) univariate Cox regression analysis of the identified characteristic genes to select prognosis-related genes; (b) LASSO regression was used for dimensionality reduction; (c) multivariate Cox regression analysis identified three genes MAS1L, TCP10L2, and CRHR2 for model construction; (d) overview of the riskscore in the training model; (e) KM survival curve of high- and low-risk patients in training cohort; (f–h) ROC curves of 1-, 3-, and 5-years survival in the training cohort; (i) overview of the riskscore in the validation model; (j) KM survival curve of high- and low-risk patients in validation cohort; (k–m) ROC curves of 1-, 3-, and 5-years survival in the validation cohort.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

(i)

(j)

(k)

(l)

(m)