In this webinar, Dr. Victoria Catterson (VP of Data Science Research at Biosymetrics Inc (a Renovaro company)) shares how to build a machine learning pipeline to uncover subtypes of Parkinson’s disease within the PPMI dataset, detailing the steps taken to clean, preprocess, and engineer features from various clinical PPMI data tables, with tips and tricks to maximize successful discovery. Dr. Catterson's talk also covers different clustering methods and the various metrics used to determine when a ‘good fit’ has been found, as well as the main clinical findings derived from this approach. This webinar was organized by the the Michael J. Fox Foundation's Data Community of Practice (DCoP). Do you have ideas or suggestions for other webinar topics you would like to see? Is there a tool you feel the community would benefit from highlighting? Let us know by leaving your thoughts in this thread: Seeking Webinar Ideas and Requests from the Community, or by contacting researchcommunity@michaeljfox.org. For those interested in joining or contributing to the DCoP, please visit rcop.michaeljfox.org.