PROJECT SUMMARY Parkinsonism, the second most common movement disorder, affects over 900,000 Americans and its prevalence is rising, while dystonia, the third most common, afflicts 250,000 Americans. X-linked dystonia parkinsonism (XDP) is a rare neurogenetic movement disorder with a wide phenotypic spectrum ranging from a parkinsonism indistinguishable from Parkinson's disease (PD), generalized dystonia, similar to DYT1 (a hereditary dystonia), or combined dystonia and parkinsonism. Female XDP carriers are at risk for parkinsonism, given X-inactivation and represents a possible genetic risk factor. As such, XDP serves as an excellent model of both dystonia and parkinsonism, and insights can inform both phenotypes and mixed movement disorders, which are notoriously challenging to assess. To provide adequate clinical trial endpoints, rater-independent, quantitative assessments of motor function are urgently needed to identify early abnormalities which may predate overt clinical symptoms and track disease progression, given the unreliability of rater-dependent clinical rating scales. This project will provide Dr. Stephen with a skill set that will allow him to assess potential quantitative measures of disease severity and progression in dystonia and parkinsonism using motion sensors and compare these measures with proposed biochemical biomarkers, using machine learning to define optimal parameters. This project aims to address three key knowledge gaps in dystonia and parkinsonism, using pure phenotypes (DYT1 and PD) and in combination (XDP): Aim 1) to utilize technology-based evaluations as more sensitive and accurate measures of dystonia in isolation (DYT1) vs. in combination with parkinsonism (XDP) compared to clinical scales; Aim 2) to examine the accuracy of sensor-based monitoring of disease progression over 1 year in XDP, PD and DYT1 patients; and Aim 3) to analyze 2 proposed blood biomarkers (one specific to XDP, and neurofilament light chain, a general marker of neurodegeneration) in XDP and combine these motor and blood markers, using machine learning to predict disease phenotype/biotype and clinical course. This research complements the NINDS objective of clinical trial readiness in rare neurological disorders, with a wider goal of better understanding these common phenotypes in the context of a mixed movement disorder, using innovative technology and analysis methods. The goal of this award is to prepare the candidate to become a fully independent investigator in the quantitative assessment of dystonia, parkinsonism and other movement disorders, in the setting of expert mentorship. The career development plan includes training goals: 1) learning transferrable motion analysis skills, 2) learning machine learning techniques to develop predictive models of motor behaviors and combining these with blood biomarker data; and 3) an introduction to biomarker science and bioinformatics. Successful completion of this project will put Dr. Stephen i...