Bioinformatics Platform Allowing Data Aggregation Across Projects and Repositories In Autism Spectrum Disorder Research: Interactive Demonstration

Background: The National Database for Autism Research (NDAR) was designed to help the ASD research community to accelerate discoveries by facilitating scientific collaboration, communication, and sharing of detailed research data. NDAR supports all types of research data, including clinical assessment, imaging and genomics, and enables researchers to search and aggregate data from multiple projects across multiple data repositories. Meaningful sharing, reaggregation, and reanalysis of data are made possible by having investigators define their data via the NDAR Data Dictionary and by their use of the Global Unique Identifiers (GUIDs). Such unifying approaches, coupled with cutting-edge grid architecture and an active partnership with other major public and private ASD-relevant informatics resources, are providing the foundation for a common informatics environment for the ASD research community, and may transform the paradigm of ASD research to one of community science. 

Objectives:   The objectives of this unscripted interactive demonstration are: to illustrate NDAR’s user-friendly interface and useful built-in data standardization, organization and validation tools; to demonstrate the ease with which data can be submitted to NDAR, requiring only minimal adjustments to the way data are collected; to update IMFAR 2011 attendees on the progress of establishing data federation with other informatics resources (e.g.,  IAN, AGRE, dbGaP); and to emphasize the power of federating available public and private ASD-related resources and repositories by demonstrating how a single query in NDAR can yield results across all repositories.

Methods: NDAR’s tools define data structures in a clear and simple way; they were developed after thorough analyses of the needs of the ASD investigators.  The NDAR Data Dictionary currently defines over 25,000 variables for clinical assessments, imaging, and genomics; it allows researchers to define their own data structures and operates with NDAR’s validation tool to ensure data quality and standardization of data. NDAR’s Genomics Tool standardizes the naming of data processing and analysis protocols; it requires entering sufficient details and enforces unambiguous interpretation of the entered information. NDAR’s global unique identifiers (GUID) protect the identity of research subjects, while allowing for the collection and analyses of data across time and projects. 

Results:  To date, ASD researchers have submitted and shared data from more than 12,000 subjects. Many others are in the process of being shared, and NDAR has as its goal to make available data from 90% of all human subject studies that commence in 2012. Launched in December of 2010, the NDAR Genomics Tool will be used by ASD investigators to define their genomics data allowing data reaggregation across projects and repositories.  NDAR not only continues to develop its methodology and capabilities, but is also working closely with the research community and other major private and public informatics efforts to form a rich global network of data and tools.

Conclusions: This is a demonstration how NDAR will add value to ASD research beyond the sum of the contributors of the individual projects and platforms, giving researchers access to more data than any one researcher or any one lab could collect, along with access to a range of robust analytical tools.