Research Paper Advance Articles

An interactive nomogram based on clinical and molecular signatures to predict prognosis in multiple myeloma patients

The validation of the Lasso model using testing and the whole dataset of GSE24080. (A, B) Survival probabilities (Prob-min and Prob-lse) were predicted by two Lasso models based on two ideal parameters Log λ.lse and Log λ.min using testing and the whole dataset of GSE24080. The Wilcoxon test was used to compare the different survival outcomes. (C, D) ROC curves analysis and the values of AUC were used to compare two Lasso models based on Log λ.lse and Log λ.min using testing and the whole dataset of GSE24080.

Figure 4. The validation of the Lasso model using testing and the whole dataset of GSE24080. (A, B) Survival probabilities (Prob-min and Prob-lse) were predicted by two Lasso models based on two ideal parameters Log λ.lse and Log λ.min using testing and the whole dataset of GSE24080. The Wilcoxon test was used to compare the different survival outcomes. (C, D) ROC curves analysis and the values of AUC were used to compare two Lasso models based on Log λ.lse and Log λ.min using testing and the whole dataset of GSE24080.