Analyzing Data from the Parkinsonís Progression Markers Initiative to Identify Subtypes of Parkinsonís Disease
Computational Science 2017, 2017
Parkinson's disease (PD) presents and progresses variably, so no two people with Parkinson's have the same disease. The diversity of PD clinical manifestations makes the disease more challenging to diagnose and treat. The goal of this project is to discover Parkinson's subtypes -- patterns of disease progression -- by analyzing clinical, neuroimaging and genetic information from people with PD.
We hypothesize that individuals with the same subtype will have similar patterns of disease progression. (We will take information about disease progression from participants' medical records.)
We will examine clinical, neuroimaging and genetic information collected in the Parkinson's Progression Markers Initiative (PPMI), MJFF's landmark study launched in 2010 to find ways to predict, track or treat PD. (There currently are no known biomarkers for Parkinson's disease.) We will use the most powerful methods of computer analysis (i.e., machine learning methods) to group similar patients together. Each group will represent a PD subtype. We then will be compare these new subtypes with other, previously identified PD subtypes (e.g., motor, cognitive, or mood).
Impact on Diagnosis/Treatment of Parkinson's disease:
Knowing PD subtypes can help us better understand how Parkinson's progresses differently from person to person, and over time. These subtypes can help inform therapeutic development for each of the stages of PD.
Next Steps for Development:
Researchers next will need to confirm the identified Parkinson's subtypes in larger groups of people with PD. (PPMI currently has nearly 1,000 participants.)
Assistant Professor at Weill Cornell Medicine
Location: New York, New York, United States