Machine learning identifies lactate metabolism biomarkers and deciphers immune infiltration landscapes in Parkinson's disease.

Parkinson's disease (PD), the second most common neurodegenerative disorder worldwide, is characterized by irreversible neuronal loss. Its progressive motor and non-motor symptoms-including resting tremor, postural instability, and autonomic dysfunction-substantially impair quality of life and impose significant socioeconomic burdens. Current diagnosis relies primarily on subjective motor assessments due to the absence of objective biomarkers, often delaying detection until advanced stages of neurodegeneration. Therefore, identifying PD-specific biomarkers is critical for early diagnosis, targeted interventions, and disease management. Recent evidence suggests a strong association between PD and dysregulated lactate metabolism, involving altered lactate levels, lactylation-dependent epigenetic modifications, and abnormal expression of genes related to lactate metabolism, all contributing to disease pathogenesis. To identify novel biomarkers associated with lactate metabolism for early PD detection, we integrated two GEO datasets as training cohorts. Differential expression analysis identified 119 differentially expressed genes (DEGs), of which 33 overlapped with 5,193 lactate metabolism-related genes. Using machine learning algorithms, including the least absolute shrinkage and selection operator (LASSO) and support vector machine (SVM), we prioritized four candidate biomarkers: KCNJ6, PDK4, LRP2, and DENR. External validation confirmed their diagnostic utility. A nomogram model constructed with these markers demonstrated robust predictive performance for PD. Immune infiltration analysis further revealed distinct immune cell profiles in the substantia nigra of PD patients and their significant correlations with the identified biomarkers. Collectively, these genes associated with lactate metabolism represent promising diagnostic indicators, offering valuable insights for both PD research and clinical application.