Lysosomal polygenic risk score in Parkinson's disease across populations.

Background: Lysosomal dysfunction has been implicated in Parkinson's disease (PD) pathogenesis. However, lysosomal polygenic risk scores (lyso-PRS) have not been evaluated extensively in PD across populations. Objectives: (1) Evaluation of lyso-PRS (lyso-PRS) performance for PD risk across populations; (2) Comparison of lyso-PRS across idiopathic PD (iPD), GBA1-PD, non-manifesting GBA1 carriers (NMC), and healthy controls (HC) in the European population. (3) Assessment of the lyso-PRS association with clinical outcomes in the PD patients of European population. Methods: We analyzed Neurobooster array data from the Global Parkinson's Genetics Program (GP2) dataset, including 32,054 iPD cases and 18,188 healthy controls across ten ancestries. Results: (1) lyso-PRS performed best in Europeans (European-GP2 test cohort: AUC = 0.602, p = 2.02e-172; European-TUEPAC validation cohort: AUC = 0.632, p = 5.34e-20) and worse in other populations (AUC: 0.51-0.59, p<0.05). (2) In the European population, the lyso-PRS distinguished GBA1-PD from HC (AUC = 0.621, p = 7.83e-71) and from NMC (AUC = 0.621, p = 1.05e-18). (3) In the European-GP2 cohort, higher lyso-PRS was significantly associated with earlier age at onset ({beta}=-0.21, p=0.022) and higher Unified Parkinson's Disease Rating Scale (UPDRS) scores ({beta}=0.32-1.13, all p < 0.05). Conclusions lyso-PRS was associated with iPD and GBA1-PD status, with the strongest predictive performance in the European population, underscoring the need to improve accuracy in other ancestries. In the European population, the lyso-PRS showed associations with clinical features; however, not all findings were replicated in the validation cohort, highlighting the need for further investigation in large and diverse cohorts.