Fig. 2.

The feature selection process. a LASSO coefficient profiles of the 841 features. b Tuning parameters (λ) selected in the LASSO model were used for applied 10-fold cross-validation with the minimum criteria. The Y-axis indicates misclassification errors, and the lower X-axis indicates the average deviance ln(λ) values, which were − 2.19, − 2.44, − 2.14, − 1.61, − 1.75, and − 2.61 in models 1–6, respectively. The vertical lines through the red dots show the upper and lower limits of the deviances. Dotted vertical lines were drawn at the optimal values using the minimum criteria with 1 standard error (the 1-SE criteria). Numbers along the upper X-axis represent the average number of predictors