FIG 2.

The dimensionality reduction results of HPV16 E6 mutation features by LASSO. (A, B) The coefficients in the Lasso regression for E6 mutation features screening. (A) For determining the adjustment parameter (λ value), we selected the adjustment parameter (λ value) with the minimum value of the misclassification error using the cross-validation method. (B) The adjustment parameter (λ value) was used to obtain the fraction deviance explained value. The obtained fraction deviance explained value was used to determine the final model selection. (C) Different mutation features among HSIL and NHSIL patients by hierarchical clustering.