Fig. 4

Random forest model showing the top 10 variables for distinguishing patients with GD from controls. A database 1 and B database 2. Gini impurity describes the probability of incorrect classification at a given node in the decision tree based on training data. Mean decrease in Gini is a measure of a variable’s importance for determining GD across all the decision trees in the random forest: the mean of a variable’s total decrease in node impurity, weighted by the proportion of samples reaching that node in each decision tree in the random forest. *HDL cholesterol < 35 mg/dL. Bone_dx, bone issues, including pain, crises, avascular necrosis, and fractures; HDL high-density lipoprotein