Scientists Mapped IBD Genetic Risk Across 2.2 Million Single Cells

If you have inflammatory bowel disease (IBD) or you're trying to figure out whether your symptoms trace back to it, you've likely hit a familiar wall: doctors can name the disease category, but not always explain why your immune system and gut lining turned on each other in the first place. A team led by researchers at the Wellcome Sanger Institute and the Open Targets consortium just published the largest attempt yet to answer that question at the level of individual cells, tracking how genetic risk variants behave inside 2.2 million cells pulled from gut and blood tissue. The result isn't a new drug or a diagnosis. It's a far more precise map of which genes, in which specific cell types, are most likely driving disease risk, the kind of map that could reshape how future IBD treatments are chosen.

Key takeaways

• A 2.2 million cell atlas of gut and blood tissue linked genetic risk to specific genes at more than half of the known IBD risk regions, nominating 419 candidate genes, 257 of them entirely new.
• Cell type specific gene activity data matched known IBD risk genes 3.5 times more often than older, whole tissue data, pointing to reduced Notch signaling in immune cells and Wnt driven MYC activity in gut stem cells as two distinct contributors to risk.
• Two existing IBD drugs, vedolizumab and tofacitinib, already target genes that turned up independently in this genetic map, and the team flagged a third gene, PRKCB, as a promising drug candidate.
• This is a genetics study, not a microbiome study. The gut bacteria angle is inference, not data, and the paper's own authors call the newly linked genes "nominated," not proven causal.

A 2.2 Million Cell Map Ties IBD Risk to Specific Genes

Researchers built a single cell genetic atlas called IBDverse, tracking how DNA variants change gene activity inside 2.2 million individual cells from blood and gut tissue, then used it to identify likely disease driving genes at more than half of the roughly 321 DNA regions already linked to inflammatory bowel disease (IBD). The team, led by Tobi Alegbe and colleagues, published the results in Nature on June 3, 2026. Most known IBD risk regions sit in "non coding" DNA, stretches that don't build proteins themselves but can turn nearby genes up or down, which is a big reason identifying the actual responsible gene at each region has been slow going for two decades of genetic research. By tracking gene activity cell type by cell type instead of averaging across a whole tissue sample, the researchers found cell type specific effects were 3.5 times more likely to line up with known IBD risk regions than older, blended tissue level measurements, and used that resolution to nominate 419 candidate effector genes, 257 of which had never been linked to IBD before.

Why This Matters for the Nearly 5 Million People Living With IBD

Anyone managing Crohn's disease or ulcerative colitis, the two main forms of IBD, has a stake in this, because current treatments are still largely built around a blunt strategy: suppress the whole immune system and hope the gut lining recovers, rather than fixing whatever specific process actually failed. The study's authors note that IBD affects more than 4.9 million people worldwide, a burden that hasn't shrunk despite a growing menu of biologic drugs. In the United States alone, the CDC recently estimated 4.1 million adults have been diagnosed with Crohn's disease or ulcerative colitis, up from 3.1 million in 2015 to 2016. What the new gene map adds is a rationale for why some patients respond to one drug class and not another: if a given patient's IBD risk traces mainly to immune cell dysfunction, an anti-inflammatory drug might work well; if it traces instead to a gut lining that can't repair itself fast enough, suppressing the immune system further may not touch the underlying problem at all.

How Researchers Built the IBDverse Cell Atlas

The team sequenced single cell gene activity from 275 healthy rectal biopsies, 119 Crohn's disease terminal ileal biopsies, 243 healthy terminal ileal biopsies, and 95 blood samples from Crohn's patients, drawn from 421 people total, 125 of whom had IBD, complementing a companion single cell study of Crohn's ileal biopsies from the same research effort. After quality control, they sorted more than 1.8 million cells into 86 distinct cell type clusters across 9 major cell populations, then matched gene activity patterns to each person's genetic data to calculate expression quantitative trait loci, or eQTLs (a statistical link between a specific DNA variant and how strongly a nearby gene gets turned on or off), at the resolution of individual cell types rather than whole tissue samples. Their final analysis used matched genetic and expression data from 396 people, identifying 88,548 eQTLs across more than 20,000 genes. They then checked which of these cell type specific eQTLs statistically overlapped, or "colocalized," with the genetic risk regions already flagged by genome wide association studies (GWAS) of IBD, a technique that identifies a shared underlying DNA variant rather than proving one thing causes the other.

What This Could Mean If You're Managing IBD

For someone living with IBD, or investigating symptoms that might be it, this study doesn't translate into a new test or treatment to ask for today. It's a mechanism map, not a clinical tool. What it does offer is a more concrete vocabulary for a future conversation with a gastroenterologist about why a drug is or isn't working: the researchers found that two existing IBD drugs, vedolizumab and tofacitinib, target genes (ITGA4 and JAK2, respectively) that showed up directly in this genetic map, reassuring validation that the approach captures real biology rather than statistical noise. They also flagged PRKCB, a gene already in early stage clinical trials for IBD, as a promising drug target, and warned that a locus near NDUFAF1, a target of the diabetes drug metformin, may help explain why metformin commonly causes gastrointestinal side effects. None of this changes what a patient should do this week. It's a preview of where drug development is heading, not a shortcut to it.

Why "Nominating" a Gene Isn't the Same as Proving It Causes Disease

The biggest limitation is built into the study's own language: the authors describe their results as "nominating" effector genes, not confirming them, because colocalization, the statistical overlap between a gene's activity pattern and a disease risk region, is strong genetic evidence of a shared underlying variant, but it is not proof that changing the gene's expression actually causes the disease. It's also worth being explicit about what this study is not. It is a genetics and gene expression study, not a microbiome study. The researchers did not sequence gut bacteria or measure the microbiome directly. The idea that a genetically weakened, slow to repair gut lining could let ordinary gut microbes exert outsized influence on disease is a mechanistic inference, raised by an accompanying commentary on the paper, not new data generated here. The study also has a real statistical soft spot: 57% of the cell types identified, including all connective tissue (stromal) cell populations, had fewer than 100 cells per person, which limits how much confidence researchers can place in rare or disease specific cell types even with 2.2 million cells in the full dataset.

What Comes Next for IBD Drug Development

The researchers frame IBDverse as a template rather than a finished product: a similar cell type resolved mapping strategy could be applied to other complex diseases where GWAS has identified risk regions but not the responsible genes. For IBD specifically, the paper points toward larger single cell studies in more people with active disease, and toward statistical methods that account for multiple variants at once, which could push discovery past the roughly half of known risk regions this study managed to explain. No timeline for clinical application is given. This is a hypothesis generating map, not a drug development pipeline.

Common Questions

Did this study look at the gut microbiome? No. It mapped how DNA variants affect gene activity in human gut and blood cells. It did not sequence gut bacteria or measure microbiome composition directly.

What is an eQTL? An expression quantitative trait locus is a statistical link between a specific DNA variant and how strongly a nearby gene gets turned on or off. This study measured these at the resolution of individual cell types rather than a whole tissue sample.

Does this study prove which genes cause IBD? No. Colocalization, the overlap between a gene's activity pattern and a known IBD genetic risk region, is strong evidence of a shared underlying variant, but the study's own authors describe the genes as "nominated," not confirmed as causal.

How many people were in the study? The researchers analyzed roughly 2.2 million single cells from 421 individuals, including 125 people with IBD, drawn from gut biopsies and blood samples.

References

1. J. L. E. Tearle, K. R. James (2026). Genetic Risk Meets the Intestinal Barrier. Immunology & Cell Biology. https://doi.org/10.1111/imcb.70158
2. Single-cell RNA sequencing of terminal ileal biopsies identifies signatures of Crohn's disease pathogenesis. Nature. https://www.nature.com/articles/s41588-026-02634-7
3. Matthew Luzentales-Simpson, Yvonne C. F. Pang, Ada Zhang, James A. Sousa, Laura M. Sly (2021). Vedolizumab: Potential Mechanisms of Action for Reducing Pathological Inflammation in Inflammatory Bowel Diseases. Frontiers in Cell and Developmental Biology. https://doi.org/10.3389/fcell.2021.612830