Two Big Data to Knowledge (BD2K) Centers of Excellence, the Center for Causal Discovery (CCD) in Pittsburgh and the Patient-Centered Informatics Common: Standardized Unification of Research Elements (PIC-SURE) at Harvard Medical School, have been awarded a supplement of $653,000 from the National Institutes of Health to prototype a federated data ecosystem that shares both big data and data analytic services in a secure and scalable manner in the cloud.
CCD investigator Jeremy Espino, MD, MS and PIC-SURE investigator Paul Avillach, MD, PhD will co-supervise this project under the leadership of Gregory Cooper, MD, PhD (CCD PI) and Isaac Kohane, MD, PhD (PIC-SURE PI).
The software being developed will provide secure cloud-based RESTful APIs for accessing a large autism dataset. Causal discovery algorithms, which are also deployed to the cloud, will access this data to construct probabilistic graphical models, which can be displayed using visualization tools.
The project, which should be completed by the end of 2016, serves three main purposes: it provides a proof-of-principle model for sharing and analyzing biomedical data in a secure and scalable manner in a “commons” area in the cloud, it may discover new knowledge about autism, and it serves as the foundation for future collaborations between the CCD and PIC-SURE.