AD Workbench: Secure, Cloud-Based ADRD Data Sharing, Analysis and Interoperability Layer
Fragmented data access, inconsistent harmonization standards, and limited infrastructure for cross-institutional collaboration have constrained Alzheimer's disease and related dementias (ADRD) research for decades. AD Workbench, the Alzheimer's Disease Data Initiative's secure, cloud-based research environment for neurodegenerative disease data sharing and analysis, addresses these barriers directly. Developed with support from a coalition of leading academic, governmental, philanthropic, and industry organizations and powered by the Global Research and Imaging Platform’s (GRIP), AD Workbench gives researchers worldwide a single environment to discover, request, access, and analyze ADRD data across institutions and platforms. More than 10,000 researchers from institutions worldwide have registered, and registration is open to qualified researchers at academic institutions and research organizations at no cost.
What Can Researchers Do in AD Workbench?
Researchers using AD Workbench can:
- Request access to 340+ curated datasets from across the ADRD research community, spanning longitudinal clinical cohorts, neuroimaging studies, multiomics collections, and biofluid biomarker datasets representing 5.8 million research participants in total.
- Analyze data within secure, permissioned workspaces equipped with virtual machines running Linux or Windows, with Jupyter Notebooks and R Studio built in. The Data Science Template includes Docker, GitLab, Julia, LibreOffice, and Visual Studio Code pre-installed. R Shiny apps are supported. Standard workspace VMs include 4 vCPUs and 16 GB of memory, with larger configurations available on request.
- Bring approved external datasets, custom code, and proprietary models into a workspace environment, enabling researchers to work with their own materials alongside platform-hosted data. Free compute is included.
- Collaborate across institutions within shared, permissioned workspaces, supporting multi-site studies and consortium-scale analysis without requiring data movement or additional infrastructure. Analytical results can be exported safely via an airlock feature that enforces data contributor agreement requirements.
- Explore available datasets through the AD Discovery Portal, a publicly accessible dataset catalog for Alzheimer's disease and related dementias research built by the AD Data Initiative, prior to submitting a formal data access request.
What Datasets Are Available Through AD Workbench?
AD Workbench provides access to 340+ curated datasets spanning clinical assessments, structural MRI and amyloid PET neuroimaging, genomics, proteomics, transcriptomics, and digital biomarkers. Datasets originate from established ADRD research programs and contributing institutions across academia and industry and are made accessible in compliance with each contributor's data governance and consent frameworks.
Datasets are built on FAIR principles; those ensuring data are Findable, Accessible, Interoperable, and Reusable. Contributions come from academia, industry, government, and nonprofit organizations, including datasets from other interoperable data-sharing platforms. Access to individual datasets may be centralized, distributed, or federated, accommodating a range of data security and governance requirements while maximizing permissioned access.
One example of available datasets is the Global Neurodegeneration Proteomics Consortium (GNPC) v1 dataset, available through AD Workbench, illustrates the scale of what the platform can support: a harmonized proteomics resource spanning 250 million unique protein measurements from 35,000 biofluid samples, contributed by 23+ data providers, with more than 100 concurrent researchers working in the platform.¹
How Does AD Workbench Support Open Science?
The Alzheimer's Disease Data Initiative was established to increase data sharing, ease data access, and develop new tools and analytics for ADRD researchers globally. AD Workbench is the operational center of that mandate. Researchers can contribute datasets to the platform through a structured curation and submission workflow, extending the platform's value for the broader field. The platform is built on Microsoft Azure and powered by GRIP, which holds ISO/IEC 27001 certification, meets GDPR and HIPAA compliance requirements, and has a product offering that meets NHS data security standards. Data hubs in Europe and the United Kingdom support GDPR data storage requirements.
Access is geographically equitable by design. Researchers at institutions across income levels, including those in low- and middle-income countries, have the same access to compute resources, tools, and datasets as researchers at high-income institutions.
How Do I Get Started with AD Workbench?
Getting started requires registration, workspace setup, and a dataset access request. All three steps are completed within the platform.
- REGISTER ↗ for an account an AD Workbench Account
- Develop your research plans by exploring datasets and metadata in the DISCOVERY PORTAL CATALOGUE ↗
- REQUEST A WORKSPACE ↗, the secure container environment in which datasets are accessed and analyzed.
- Submit a DATASET ACCESS REQUEST ↗ dataset access request, specifying the workspace to which approved data will be provisioned.
- Once the data access request is approved, ACCESS DATA ↗ within the designated workspace and analysis can begin.
Technical documentation for each step is available in the AD Workbench Tech Help section, including the Workspace Quick Start Guide, dataset request procedures, and workspace administration reference.
Frequently Asked Questions
What is AD Workbench?
AD Workbench is the Alzheimer's Disease Data Initiative's secure, cloud-based research environment and interoperability layer for ADRD data sharing and analysis. It provides qualified researchers with access to 340+ curated datasets, collaborative permissioned workspaces, and multimodal analysis tools. AD Workbench brings together 10+ data platforms supported by organizations like Vivli, Sage Bionetworks, Alzheimer's Association, Medical Research Council UK, American Heart Association & the NIA. It is a purpose-built ADRD research platform and is not affiliated with similarly named commercial database tools.
How is AD Workbench different from the AD Discovery Portal?
The AD Discovery Portal is a publicly accessible dataset catalog where any researcher can browse available datasets, review study details, and initiate an access request without registering first. AD Workbench is a secure research environment where approved datasets are accessed, analyzed, and combined. The two are fully integrated: dataset requests initiated in the AD Discovery Portal are provisioned directly to a researcher's AD Workbench workspace once approved.
Can I bring my own data and code into AD Workbench?
Yes. Researchers can upload approved external datasets, import custom analysis code, and run proprietary models within their workspace environment. AD Workbench workspaces support R and Python runtimes and include virtual machines with free compute, enabling researchers to work with both platform-hosted and researcher-supplied materials in the same environment.
What compute resources are available in AD Workbench workspaces?
Standard AD Workbench workspaces include virtual machines with 4 vCPUs and 16 GB of memory, with larger configurations available based on research plan or on request. Multiple VMs can be added to a single workspace, sharing a common storage pool of up to 5 TB. Jupyter Notebooks and R Studio are built-in workspace features. The Data Science Template includes Docker, GitLab, Julia, LibreOffice, and Visual Studio Code pre-installed, and additional software can be installed on Linux or Windows VMs. Full documentation is available in the AD Workbench Tech Help section.
What other tools does the Alzheimer's Disease Data Initiative offer?
AD Workbench is part of a broader suite of research infrastructure developed by the Alzheimer's Disease Data Initiative. The AD Curation Studio supports harmonization and preparation of ADRD datasets for sharing. The Federated and Distributed Data Sharing Appliance (FDSA) enables federated data analysis across institutional boundaries without requiring data movement. The AD Data Initiative Repository supports long-term data stewardship. Descriptions of each product are available on the AD Data Initiative product suite overview page.
Footnotes
¹ Alzheimer's Disease Data Initiative. "AD Workbench Powers the World's Largest Protein Biomarker Discovery Effort: The Global Neurodegeneration Proteomics Consortium (GNPC)." alzheimersdata.org, 2025. https://www.alzheimersdata.org/case-studies/ad-workbench-powers-the-world-s-largest-protein-biomarker-discovery-effort-the-global-neurodegeneration-proteomics-consortium-gnpc