| Abstract: |
| QSP models take significant time and effort to create when compared to statistical or classical empirically driven models; mostly because of the length of time to identify, evaluate and populate credible model priors. We are constructing a digital research environment (DRE) that receives data from a community-centric crowdsourcing approach that includes patients (patient swarm) so that model priors can be more efficiently catalogued and evaluated by data curators and QSP modelers. Crowdsourcing efforts will be evaluated as the model is being constructed and the patient swarm will be engaged to comment on both the structure and its predictive potential to explain disease progression and evaluate historical and current development candidates in real-time. In addition, patient-generated disease trajectories will be used as a real-world data source to validate the model. Upon completion these will serve to verify that the model is able to generate synthetic data that more closely mimics the heterogeneity of the family of disease etiologies currently classified as Parkinson`s Disease. Patient research participants will describe their illness in quantitative terms with the help of an experienced QSP modeling team, some of whom will construct a model based on priors collected from all available sources (public and private sector) using an AI/ML driven text mining approach to identify source data from the literature. |
|