fig1

Figure 1. Feature selection using LASSO regression and bootstrap stability analysis. (A) LASSO coefficient paths for the candidate predictors; (B) Cross-validation error curve for selecting the optimal penalty parameter λ; the dotted lines indicate λ_min and λ_1se; (C) Selection frequency of each predictor across 100 bootstrap LASSO iterations; the red dashed line denotes the 90% threshold for final retention. LASSO: Least absolute shrinkage and selection operator; λ_min: minimum penalty parameter; λ_1se: largest penalty parameter within one standard error of the minimum cross-validation error; WBC: white blood cell count; MCV: mean corpuscular volume; MET: metabolic equivalent of task; TG: triglycerides; HbA1c: glycated hemoglobin (hemoglobin A1c); SBP: systolic blood pressure; HDL: high-density lipoprotein; PLT: platelet count; DBP: diastolic blood pressure; CRP: C-reactive protein; BMI: body mass index; FBG: fasting blood glucose; UA: uric acid; BUN: blood urea nitrogen; LDL: low-density lipoprotein; TC: total cholesterol.