Urine biomarkers discovery by metabolomics and machine learning for Parkinson's disease diagnoses

Parkinson's disease (PD) is a complex neurological disorder that typically worsens with age. A wide range of pathologies makes PD a very heterogeneous condition, and there are currently no reliable diagnostic tests for this disease. The application of metabolomics to the study of PD has the potential to identify disease biomarkers through the systematic evaluation of metabolites. In this study, urine metabolic profiles of 215 urine samples from 104 PD patients and 111 healthy individuals were assessed based on liquid chromatography-mass spectrometry. The urine metabolic profile was first evaluated with partial least-squares discriminant analysis, and then we integrated the metabolomic data with ensemble machine learning techniques using the voting strategy to achieve better predictive performance. A combination of 8-metabolite predictive panel performed well with an accuracy of over 90.7%. Compared to control subjects, PD patients had higher levels of 3-methoxytyramine, N-acetyl-l-tyrosine, orotic acid, uric acid, vanillic acid, and xanthine, and lower levels of 3,3-dimethylglutaric acid and imidazolelactic acid in their urine. The multi-metabolite prediction model developed in this study can serve as an initial point for future clinical studies.