Fig. 2

Prediction modeling of gout patients with urate renal underexcretion (RUE). A The area under the receiver-operator characteristic curve (AUC) of different numbers of 73 variables (42 SNP variations and 31 clinical parameters) revealed by the LASSO model in the derivation set. The red dots represent the AUC score, the gray lines represent the standard error, and the vertical dotted lines represent optimal values by minimum criteria. The upper abscissa is the number of non-zero coefficients in the model at this time, the lower abscissa is log λ, which is the tuning parameter used for 10-fold cross-validation in the LASSO model. A dotted vertical line is drawn at the optimal values by minimum criteria, which is 11. B LASSO coefficient profiles of the 73 variables. A vertical line is drawn at the optimal value by 1−SE criteria and results in 11 non-zero coefficients. C The receiver-operator characteristic analyses for predicting RUE in the internal test set with stochastic gradient descent. D The precision-recall curve of predicting RUE in the internal test set. E The receiver-operator characteristic analyses for predicting RUE in the validation set with stochastic gradient descent. F The precision-recall curve of predicting RUE in the validation set