Fig. 7

Development of a five-CpG-sites-based prognostic prediction model of pancreatic cancer using Lasso regression. a The λ value of the lasso model was decided by cross-validation with minimal misclassification. b Five CpG sites and their corresponding coefficients were utilised to construct the prognostic risk model. c Distribution plot of the risk score and survival time between high- and low-risk groups in the training set. d Heat map of the five CpG sites in the training set. e Overall survival analysis between the high- and low-risk groups in the training set. f Distribution plot of the risk score and survival time between high- and low-risk groups in the validation set. g Heat map of the five CpG sites in the validation set. h Overall survival analysis between the high- and low-risk groups in the validation set