NIH 'All of Us' Program Releases World's Largest Integrated Genomic and EHR Dataset

On June 30, 2026, the National Institutes of Health (NIH) issued the most expansive data release in the history of its All of Us Research Program, fundamentally altering the landscape of biomedical research. The release, known as CDRv9, establishes the program as the world's largest integrated database of genomic and electronic health record (EHR) information. By combining unprecedented genomic depth with real-world clinical breadth, the NIH has handed scientists a comprehensive map of human health designed to accelerate the transition toward precision medicine.[1][3][4][5]

The sheer scale of the new dataset represents a logistical and scientific milestone. The release includes data from more than 747,000 participants, enabling the direct linkage of nearly 482,000 electronic health records to 535,000 whole genome sequences. In total, the repository now houses more than 1.3 billion genetic variants, alongside 553,000 genotyping arrays and roughly 600,000 physical measurements. This volume of data provides researchers with the statistical power necessary to untangle the complex web of factors that dictate human health.[1][2][3][5]

The necessity of this scale is rooted in what NIH Director Jay Bhattacharya described as the central paradox of precision medicine. To successfully tailor medical treatments to a single individual, researchers must first analyze massive populations to uncover the subtle patterns connecting genetics, lifestyle, and environmental exposures to specific health outcomes. Without a cohort of hundreds of thousands of people, the statistical signals of rare diseases and nuanced drug interactions remain buried in the noise.[1][3]

What truly separates the All of Us dataset from its predecessors is not just its size, but its composition. More than 645,000 participants—representing 86% of the total cohort—come from communities that have been historically underrepresented in biomedical research. This includes racial and ethnic minorities, rural residents, older adults, and people with disabilities, spanning all 50 U.S. states and covering more than 98% of three-digit ZIP codes.[1][3][4]

This diversity is a deliberate corrective to legacy genomic databases. Historically, repositories like the UK Biobank—which holds records for roughly 500,000 individuals—have been overwhelmingly composed of people of white European ancestry. Because genetic variants can differ significantly across populations, findings derived from homogenous datasets often fail to translate to other groups, leading to diagnostic tests and medications that are less effective for minorities.[4][7]

The program's inclusive approach is already yielding dividends. Previous analyses of the All of Us cohort have identified over 275 million genetic variants that had never been reported in prior large-scale projects. By capturing this hidden genetic variation, researchers can begin to close the translational gap, ensuring that the next generation of targeted therapies and risk-prediction models are equitable and effective for all populations.[4][7]

However, a sequenced genome is only a static blueprint; its true value is unlocked when paired with clinical outcomes. The linkage of 535,000 genomes to 482,000 EHRs bridges the gap between genetic codes and real-world diagnoses. This allows scientists to observe how specific genetic mutations manifest as diseases over time, how quickly those diseases progress, and why different patients experience varying responses or adverse events to the exact same medication.[2][5]

The dataset extends far beyond the walls of the clinic. In addition to medical records, researchers have access to 747,000 survey responses capturing social circumstances and behaviors, as well as data from wearable devices like Fitbits. This holistic approach acknowledges that genetics alone cannot explain most health outcomes, allowing investigators to evaluate the combined effects of biology, socioeconomic factors, and local environments, such as air quality.[1][2][4]

The June 2026 release also marks a critical technological pivot: the program's entry into the "multiomics" era. While whole genome sequencing maps the DNA, multiomics adds layers of functional data to show how that DNA is actively being used by the body. This provides unmatched statistical power and a more dynamic view of human biology.[1][5]

For the first time, the dataset includes proteomics data—the study of proteins produced by the body—from nearly 10,000 participants. It also incorporates RNA sequencing (RNA-seq) data from nearly 9,000 participants, which reveals how genes are turned on or off in different cells. Together, these modalities allow researchers to trace the entire biological pathway from a genetic mutation to the specific protein malfunction that causes a disease.[1][5]

Furthermore, the release introduces long-read whole genome sequences from more than 14,500 participants. Traditional short-read sequencing chops DNA into tiny fragments, which can make it difficult to map highly repetitive or complex regions of the genome. Long-read technology sequences much larger continuous strands, providing the resolution needed to detect large structural variations that are often implicated in rare genetic disorders.[1][5]

The utility of the All of Us platform is already well-documented. To date, nearly 23,000 registered researchers from institutions across the globe have utilized the cloud-based Researcher Workbench. This widespread access—which provides scientists at rural universities the same computational tools as those at major research hubs—has fueled the publication of more than 1,400 peer-reviewed papers.[1][3]

These publications are translating raw data into tangible clinical tools. Recent breakthroughs enabled by the dataset include the validation of a low-cost prostate cancer risk model that is currently being tested in a clinical trial of 5,000 U.S. veterans. Researchers have also developed a first-of-its-kind clinical genetic test capable of predicting inherited risk across eight different cardiovascular conditions, and identified novel genetic changes linked to Alzheimer's disease prevention.[1]

Despite these successes, clinical trial sponsors and data scientists caution that significant infrastructure challenges remain. While the ability to generate genomic data at scale is largely solved, the persistent difficulty lies in standardizing the clinical data. Health records are notoriously fragmented, messy, and siloed across different hospital systems, making it difficult to turn participant consent into seamlessly linked, usable data.[3]

The NIH is actively working to address this bottleneck. In the latest release, EHR data grew by 22%, driven by strategic expansions in data sourcing. This growth was facilitated by integrating participant-mediated EHR submissions and tapping into broader health information exchange networks, slowly building a more cohesive and standardized clinical picture.[1][3]

Housing half a million genomes linked to medical records creates one of the most sensitive datasets ever assembled, placing a premium on privacy and security. The data is not publicly downloadable; it is strictly confined to the Researcher Workbench, a secure environment where verified scientists must agree to a rigorous code of conduct. Personal identifiers, such as names and addresses, are stripped from the records before researchers can access them.[4][5]

Ultimately, the foundation of the All of Us program is not its sequencing technology, but the trust of its participants. As artificial intelligence makes it increasingly easy to draw complex inferences from anonymized health data, guarding that trust will become both more difficult and more essential. The program's continued success relies on maintaining transparent communication and robust security protocols.[1][4]

With total enrollment now exceeding 883,000 individuals, the NIH is rapidly closing in on its ultimate goal of one million participants. As the dataset continues to grow in both size and multiomic depth, it cements its status as a national treasure—a foundational platform that will drive the discovery of individualized treatments and equitable healthcare solutions for decades to come.[1][4][5]