Synthesizing biomedical data into effective discoveries is greatly hampered by the way science is both reported in research articles and stored in databases. Articles do not present raw data in an accessible or understandable form and scientific databases are highly specialized, designed only to store one kind of data. Our objective is to develop generic, next-generation scientific database technology that permits storage, retrieval and synthesis of any type of data from any type of experiment.
As a proof-of-concept, we will construct a repository to capture the design of a small subset of the experiments currently funded by the Michael J Fox Foundation. We will then populate the system with data from those experiments, ensuring that our representation allows subsequent data reporting, retrieval and reuse. We will develop community-driven standard terminologies to permit comparisons across experiments. This will be based on an existing, open-source, federally-funded database system (called 'Yogo') and will involve the creation of additional tools to allow non-computer scientists to construct a viable data repository based only on a relatively simple knowledge engineering model of experimental design. Security of the system will be maintained at the maximum level possible to protect the intellectual property of the underlying data.
Relevance to Diagnosis/Treatment of Parkinson’s Disease:
The Michael J Fox Foundation provides millions of dollars for Parkinson's Research but currently has no centralized repository to bring together all the data from the scientific experiments it supports. Such a centralized database would stimulate cross-linking between projects, accelerate research, and greatly improve communication between the Foundation and its grantees. Improvements in our ability to synthesize data across studies greatly improve efficiency and speed us to finding a cure.
This work will generate a viability study describing how the system performs at capturing the design of the limited set of scientific experiments under study as well as a practical, working demonstration prototype (although not a production level system, this will be a viable, working demonstration). Our viability study will include careful evaluations of the system's ease-of-use, validity and performance. All software will be open-source, available for download and use by our colleagues.