Hello, everyone!
As part of my project in the Data Modality and Methodology Task Force, I aimed to create and share Python scripts that generate clinician- and research-relevant data based on PPMI data. The project is currently in its beta version and is publicly shared with everyone via the Research Community’s GitHub.
The currently available notebooks include:
- LEDD and medication-specific LEDD calculations
- Levodopa challenge responsiveness
- Medical conditions identifier
- Medication usage identifier
Each notebook includes a theoretical background and references to research articles that justify its utility. The goal is to enable other researchers to use this data to identify meaningful clinical correlations. I believe this will be particularly useful for those with limited clinical exposure, who may sometimes struggle to identify relevant clinical outcomes to study.
For example, a researcher working with microbiome data may find it useful to check if a specific microbiome pattern or cluster is associated to the degree of levodopa responsiveness (which makes clinical sense, as disabsorption due to the intestinal bacterial profile may impair the levodopa response).
Another example is that a researcher working with neuroimaging data may try to identify some specific neuroimaging patterns that could predict longitudinal LEDD evolution. Lastly, given the relevance of the association of some medical conditions to PD, we can also use those scripts to check, for instance, if any kind of commorbidity (like diabetes) would be associated with a worsened clinical progression.
Please let me know if you find any errors in the scripts, share your thoughts on their usefulness, and suggest any additional relevant clinical data that should be included.
