Figure 4.Identification of HMs for construction of risk model. (A) Identification of differentially expressed HMs associated with prognosis of KIRC samples by univariate analysis. (B) Elimination of HMs with model overfitting by LASSO analysis. (C) Tenfold cross-validation for tuning parameter selection in the LASSO analysis.
