NIMH Data Repository Requirements

DATA RESPOSITORY

In addition to the NIH Data Management and Sharing Policy, the National Institute of Mental Health (NIMH) has the following data repository requirements for data generated by NIMH grant recipients. 

NIMH has established the NIMH Data Archive (NDA) to enable responsible sharing and use of data collected from human research participants. NDA is a cloud-based data infrastructure that provides controlled access to human subject-level data across many scientific domains to support secondary data analysis and research reproducibility. The NDA mission is to accelerate scientific research and discovery through data sharing, data harmonization, and the reporting of research results. Human subject-level data from NIMH-funded research are expected to be deposited into NDA. This includes, but is not limited to, raw and analyzed clinical, genomic, imaging, and phenotypic data.

NDA holds data collected from thousands of research projects, and provides infrastructure for sharing research data, tools, methods, and analyses enabling collaborative science and discovery.  NDA serves as a gateway to datasets from large-scale NIMH-funded research efforts including but not limited to:

• Connectome Coordination Facility
• Accelerating Medicines Partnership-Schizophrenia (AMP-SCZ) project
• Neurobiobank Data Repository and the PsychENCODE project

NDA supports not only the human subjects research funded by NIMH (located in the NDA itself), but also a number of repositories, including the Osteoarthritis Initiative (OAI), the Connectome Coordination Facility (CCF), Accelerating Medicines Partnership – Schizophrenia (AMP SCZ), NIAAA Data Archive (NIAAADA), and the Helping to End Addiction Long-term® (HEAL) Initiative. 

NDA provides access to genetic, pedigree, and clinical data associated with biospecimens available from the NIMH Repository and Genomics Resource (NRGR).

NDA Data Structures

Submitted subject-level datasets are harmonized into a common set of data structures using Global Unique Identifiers (GUIDs) or pseudoGUIDs. NDA does not hold personally identifying information (PII). NDA holds over thirty data structures populated with data from >20,000 human subjects as well as thousands of other maturing structures of varying sizes. NDA data assets are primarily categorized as clinical, -omics, or imaging data, with many subjects represented by multiple data types. 

NDA also utilizes Common Data Elements (CDEs). CDEs are data elements that have been identified and defined for use in multiple data sets across different studies. NDA currently holds over 500,000 subject-level CDE files gathered from more than 2,000 independent research projects. NIMH strongly encourages investigators to collect CDEs for all mental health human subjects research unless otherwise indicated during the negotiation of the terms and conditions of award. The funds necessary for collecting and subsequently submitting CDE data should be included in the requested application budget. A cost estimator is available to facilitate the calculation of these costs on the NDA website. 

Getting Access to Shared Data

While summary data is available to all, NDA provides controlled access to de-identified research data through Data Access Requests (DARs) which are subject to NIH-wide rules and guidance for controlled-access data repositories (CADRs; Requirements for NIH Controlled-Access Data Repositories and Users).