Comprehensive Multiome Atlas of 10 Million Immune Cells Pinpoints Genetic Drivers of Autoimmune Diseases

For years, geneticists and molecular biologists assumed that the body's most critical genes were governed by single, powerful regulatory switches. They believed that disease-causing mutations in these highly constrained regions would produce massive, easily detectable shifts in gene expression that standard tests could spot.[3]

But a massive new molecular map of the human immune system proves that assumption wrong. Published this week in Nature, the dataset reveals that our most important genes are actually buffered by dozens of weak, overlapping regulatory elements, fundamentally changing how mutations cause harm.[1]

The findings come from a population-scale atlas of more than 10 million immune cells, drawn from 1,108 Finnish blood donors. By measuring both gene expression and chromatin accessibility in the exact same cells, researchers traced how subtle DNA changes cascade into altered immune function.[1][3]

Researchers from the Broad Institute, Massachusetts General Hospital, and the University of Helsinki led the effort. Working alongside the FinnGen research project and BioBank Japan, they built a high-definition map linking tens of thousands of genetic variants to hundreds of distinct diseases.[2][3]

"Immune dysregulation sits at the root of an enormous range of human diseases, from autoimmunity to cancer to neurodegeneration," said Ramnik Xavier, a co-senior author of the study. Xavier serves as a core institute member at the Broad Institute.[2]

"What excites me about this work is that it doesn't just tell us that a gene matters; it tells us the molecular mechanism, which can point toward therapeutic strategies," Xavier added. He emphasized the immediate practical applications for future drug development.[2]

Decoding the chromatin cascade

To understand how the atlas works, researchers had to look beyond the genes themselves. Most disease-associated genetic variants do not sit inside the coding regions of genes, making their downstream effects notoriously difficult to track using traditional sequencing methods.[2]

Instead, these mutations are scattered across the vast stretches of DNA that regulate gene activity. Until now, linking a specific regulatory variant to the exact gene it controls has been a major bottleneck in precision medicine, leaving thousands of genetic risk factors unexplained.[2][3]

The new method solves this by examining chromatin accessibility—the physical openness of the DNA strand. When DNA is tightly spooled, genes are silenced; when it unwinds, regulatory elements can bind and activate gene expression, allowing the cell to produce specific proteins.[2][3]

By tracking 10 million individual cells across eight broad immune cell types, the researchers could see exactly when and where the chromatin opened. They then correlated those physical changes with the actual RNA output of nearby genes, capturing the mechanism in action.[2]

This dual-measurement approach allowed the team to trace a complete molecular chain of events. As the researchers detailed in Nature, the variants most likely to cause disease are those whose effects can be traced from open chromatin to enhancer activity to target gene expression.[1][3]

The paradox of constrained genes

The most conceptually significant discovery in the atlas resolves a long-standing paradox in evolutionary biology. Scientists have long known that certain genes are under strong evolutionary constraint, meaning they rarely mutate because changes are usually fatal to the organism.[3]

Yet, paradoxically, disease-causing variants frequently target these exact constrained genes. Previous studies struggled to explain this contradiction because they could not detect the expected large shifts in gene expression when these specific variants were present in patient samples.[3]

The new atlas reveals why, uncovering a phenomenon the authors call multilayered regulatory buffering. Instead of relying on a single master switch, these critical genes are controlled by a distributed network of weak regulatory elements that share the load.[1][3]

This buffering protects the gene from catastrophic failure if one switch breaks. However, it also means that disease risk accumulates through subtle, distributed effects that traditional single-gene studies systematically missed, hiding the true drivers of complex autoimmune conditions.[3]

To prove this mechanism, the team did not just rely on statistical models. They turned to massively parallel reporter assays, a high-throughput technique that simultaneously tests the regulatory activity of thousands of DNA sequences in the laboratory to confirm the computational predictions.[1][3]

Tracing specific autoimmune pathways

The experimental validation confirmed the statistical predictions, providing concrete mechanistic hypotheses for specific conditions. The atlas successfully traced complete regulatory cascades for several known autoimmune loci, including those driving autoimmune hypothyroidism, inflammatory bowel disease, and asthma.[2][3]

For example, the researchers mapped the pathways for TICAM1, an adaptor protein involved in Toll-like receptor signaling, and RHOH, a small enzyme critical for T cell receptor signaling. Both pathways are heavily linked to the development of autoimmune hypothyroidism.[1][3]

The team also decoded the mechanism behind a variant in the TNRC18 gene. A 2023 FinnGen study had flagged this region as a risk factor for inflammatory bowel disease, but the biological mechanism had remained a complete mystery until this multiomic analysis.[2]

The new data show that the variant reduces TNRC18 expression in T cells—most strongly in a subset called T helper 1 cells. This subtle reduction pushes those specific cells toward a more inflammatory state, actively driving the bowel disease.[2]

By openly releasing the dataset and their analytical pipelines, including a classification framework called CASCADE, the researchers have provided a foundational reference for the global scientific community. The tools are already available on GitHub for other research teams to use.[3][4]

Moving forward, the research consortium plans to expand their focus beyond healthy blood donors. The next phase of the project will involve studying immune cells extracted directly from diseased tissues, offering an even closer look at active pathology in patients.[2]

As the catalog of disease-associated genetic variants continues to grow, this atlas provides the necessary translation layer. It turns a list of genetic risks into a map of actionable biological mechanisms, paving the way for highly targeted drug development in immunology.[2][3]