The research team, led by investigators at the University of Chicago, collected pinch biopsies from the terminal ileum and the ascending colon of 23 patients with Crohn’s disease, some with active inflammation and some without, alongside 16 healthy controls. From 71 biopsies, they generated single-cell assay for transposase-accessible chromatin with sequencing, or scATAC-seq, data. After rigorous quality control, the dataset retained 178,030 individual cells, which the researchers clustered into 29 distinct cell types spanning the epithelial, immune and stromal lineages of the intestinal wall. By annotating these clusters with reference gene-expression data, they built a cell-type-resolved catalog of 557,310 candidate cis-regulatory elements, or cCREs, each representing a region of open, potentially active chromatin.

The scale of the atlas matters because Crohn’s disease is not a disease of a single cell type. It is a chronic inflammatory condition of the gastrointestinal tract that arises from an interplay of host genetics, environmental triggers and the gut microbiota. Genome-wide association studies have linked more than 320 genetic loci to inflammatory bowel disease risk, but most of those risk variants sit in noncoding regions of the genome, far from any obvious gene, and their function has remained opaque. Knowing which cell types open or close which regulatory elements provides exactly the kind of functional annotation needed to interpret those variants.

The cellular composition of the atlas reflects the biology of the gut. Epithelial cells dominated the dataset, accounting for 69.3 percent of all cells, followed by immune cells at 26.3 percent and stromal cells at 4.3 percent. Among the epithelial compartment, the researchers identified mature absorptive cells specific to the ileum, known as enterocytes, and their colonic counterparts, colonocytes, along with goblet cells, Paneth cells, enteroendocrine cells and Tuft cells. The immune compartment included CD4-positive and CD8-positive T cells, gamma-delta T cells, natural killer cells, innate lymphoid cells, plasma cells, several B cell subsets, macrophages, dendritic cells and mast cells. Notably, the team also captured neutrophils, innate immune cells that are notoriously difficult to profile with standard single-cell RNA sequencing and have consequently remained understudied despite their suspected role in inflammatory disease.

Because discrete cell-type clusters cannot fully capture dynamic processes such as inflammation or continuous differentiation, the researchers complemented their cluster analysis with a computational approach called topic modeling, which assigns each cell partial membership in multiple regulatory programs simultaneously. This revealed programs shared across related cell types as well as programs tied to specific contexts. One program, topic 7, stood out as a signature of inflammation in epithelial cells: its activity was high in biopsies taken directly from inflamed regions, moderate in adjacent noninflamed tissue from the same patients, and absent in healthy controls. The regulatory elements within this program were enriched for binding motifs of AP-1 family factors and interferon regulatory factors, implicating interferon and JAK-STAT signaling as central drivers of the epithelial inflammatory response. Putative target genes included well-known inflammatory markers such as LCN2, NOS2, DUOX2 and major histocompatibility complex class II genes.

Inflammation left its mark across the entire mucosa. Comparing inflamed tissue with healthy controls, the researchers identified inflammation-associated differentially accessible regions in 21 cell types, with epithelial cells yielding the largest number. Endothelial cells, which line blood vessels and gate the movement of cells and molecules into inflamed tissue, showed 1,702 such regions enriched for proinflammatory transcription factor motifs. CD8-positive T cells displayed accessibility changes linked to the effector regulator EOMES, plasma cells showed changes involving the maturation factor POU2F2, and macrophages exhibited broad remodeling involving AP-1, IRF and STAT family factors. In goblet cells and Paneth cell-associated programs, the data pointed to plasticity in the secretory lineage, suggesting that these mucus- and antimicrobial-producing cells reprogram their regulatory landscape in response to injury and inflammation.

To connect DNA sequence directly to chromatin state, the team exploited natural genetic variation among their 39 donors. By imputing genotypes from the sequencing data and mapping chromatin accessibility quantitative trait loci, or caQTLs, in 19 cell types, they identified thousands of genetic variants that alter chromatin accessibility in a cell-type-specific manner. Critically, they showed that variants predicted to disrupt transcription factor binding motifs correlate with reduced accessibility at the affected elements. Disruption of TCF3 motifs in plasma cells and CDX2 motifs in colonocytes were among the clearest examples, providing direct evidence that the regulatory logic inferred from motif enrichment reflects real functional relationships between sequence, protein binding and chromatin state.

The atlas then delivered its most consequential finding: a cell-type-resolved view of Crohn’s disease heritability. Using stratified linkage disequilibrium score regression on GWAS summary statistics, the researchers found that 10 of the 29 cell types showed significant enrichment of inflammatory bowel disease genetic risk, and all 10 were immune cells. Adaptive immune cells, particularly T cells, carried the strongest signal, but innate immune cells, including macrophages, dendritic cells and neutrophils, were also enriched for Crohn’s disease risk specifically. Strikingly, ulcerative colitis risk was not enriched in those innate cells, a result consistent with known differences in genetic architecture between the two major forms of inflammatory bowel disease and a demonstration of the specificity of the approach. Epithelial and stromal cells, despite their dramatic inflammatory responses, showed no enrichment for disease risk, and neither did the inflammation-associated regulatory programs themselves.

That separation between genetic risk and inflammatory consequence is one of the study’s most provocative implications. The data suggest that the genetic predisposition to Crohn’s disease operates primarily through immune cells, while the epithelial and stromal changes visible in inflamed tissue are largely downstream effects of the inflammatory process rather than primary causes. The team also linked body mass index heritability to enteroendocrine cells, confirming a previous observation and hinting that part of the genetic contribution to metabolism is mediated through these hormone-producing gut cells.

Finally, the researchers fine-mapped Crohn’s disease GWAS signals and overlaid the resulting variants onto their atlas. Variants with high posterior inclusion probabilities were 3.5-fold enriched within cCREs, and the team prioritized 30 candidate regulatory variants across 22 loci. Most fell in elements active in innate and adaptive immune cells, but several implicated epithelial contexts. Near the key risk gene NOD2, a fully prioritized variant sat in a neutrophil-associated regulatory element. Two variants near CIITA, a co-activator of MHC class II genes, occupied adjacent regulatory elements with strikingly different specificities, one active across immune and endothelial cells and the other strongly associated with epithelial inflammation, suggesting that multiple variants within a single locus can act through different cell types. The authors caution that their moderate cohort size limited power for finer clinical stratification and for detecting caQTLs in rare cell types, and larger cohorts will be needed to dissect disease subtypes and treatment responses. Even so, the atlas stands as a public resource that transforms a list of noncoding risk variants into testable hypotheses about which cells, which switches and which genes set the stage for Crohn’s disease.

Subject of Research: Single-cell chromatin accessibility mapping of the human intestinal mucosa to identify regulatory programs and cell types underlying Crohn's disease genetic risk

Article Title: Single-cell analysis of chromatin accessibility in the human intestine identifies regulatory programs and clarifies genetic associations in Crohn’s disease

Article References: Zhao, Y., Zhou, R., Mu, Z., Carbonetto, P., Zhong, X., Xie, B., Luo, K., Jiang, Z., Liu, J., Cham, C. M., Koval, J., He, X., Dahl, A. W., Liu, X., Chang, E. B., Basu, A., & Pott, S. (2026). Single-cell analysis of chromatin accessibility in the human intestine identifies regulatory programs and clarifies genetic associations in Crohn’s disease. Nature Genetics, 58(10), 2535-2549. https://doi.org/10.1038/s41588-026-02755-z

Image Credits: AI Generated

DOI: 10.1038/s41588-026-02755-z

Keywords: Crohn's disease, single-cell ATAC-seq, chromatin accessibility, regulatory elements, inflammatory bowel disease, GWAS, intestinal mucosa, T cells, neutrophils, epigenomics, fine-mapping, topic modeling

The research team, led by investigators including Xin Gao and Deqing Hu of Tianjin Medical University, assembled an enormous chromatin accessibility dataset drawn from 208 single-cell ATAC sequencing samples. These samples spanned healthy adult and fetal tissues, peripheral blood from healthy donors and patients with systemic lupus erythematosus, precancerous familial adenomatous polyposis lesions, and tumors from colorectal, clear cell renal cell, ovarian, and endometrial cancers. After computational integration and batch correction using the ArchR framework, the myeloid compartment resolved into fourteen clusters, five of which were macrophage populations with strikingly distinct tissue distributions and functional signatures.

Two of those populations stood out. Macro-SPP1 cells, defined by expression of the secreted phosphoprotein SPP1, were concentrated in cancerous tissues and showed elevated activity of genes such as IL10, PPARG, INHBA, and IL4I1, along with strong wound healing, hypoxia, and angiogenesis signatures. Macro-C1QC cells, enriched in precancerous polyp tissue, carried markers of tissue-resident macrophages such as FOLR2 and MRC1 alongside anti-inflammatory programs. Motif analysis of their open chromatin revealed candidate regulators: NF-κB-related and bZIP factors dominated the accessible regulatory regions of Macro-SPP1, whereas bHLH and C2H2 zinc finger factors marked Macro-C1QC.

To connect these accessibility patterns to gene output, the researchers integrated the chromatin data with a published pan-cancer single-cell transcriptome of myeloid cells and computed peak-to-gene links, correlations between the accessibility of regulatory elements and the expression of their target genes. Super-enhancer analysis then identified the most densely wired regulatory hubs in each population. In Macro-SPP1, the top super-enhancer-linked genes included chemokines and SOD2, and the full set of 162 high-confidence super-enhancer-regulated genes was enriched for inflammatory response, IL-10 signaling, angiogenesis, wound healing, and HIF-1 signaling. When the team scored these genes in bulk tumor data from The Cancer Genome Atlas, high signature scores predicted poor overall survival in 18 of 29 cancer types. Notably, the predicted fraction of M2-like macrophage infiltration did not track with prognosis, suggesting that the specific enhancer-driven transcriptional program, rather than the sheer number of suppressive macrophages, is what matters for patient outcomes.

Pseudotime trajectory analysis then addressed the question of where these cells come from. The results supported two divergent routes to a pro-tumorigenic identity. Macro-SPP1 cells appeared to descend from peripheral monocytes through an intermediate Mono-NLRP3 stage, while Macro-C1QC cells seemed to arise from resident Macro-LYVE1 macrophages already embedded in tissue. Along the monocyte route, early activation of inflammatory chemokines, IL1B, and THBS1 was followed later by IL10, with NF-κB and AP-1 family factors initiating the process and PPARG and NFAT5 driving the terminal state. The tissue-resident route instead involved CCL3 and CCL4 chemokines and factors such as PRDM1, HES1, ETV5, and MAF. The identification of NFAT5, a transcription factor best known for mediating adaptation to hyperosmotic stress, as a key regulator alongside PPARG and the chromatin-binding protein MECP2 was particularly intriguing, hinting that TAMs activate a broad stress-adaptation program to survive the metabolically hostile tumor microenvironment.

To test these candidates experimentally, the team used the standard THP-1 monocyte model, differentiating cells into macrophages and polarizing them toward inflammatory M1 states with interferon-gamma and lipopolysaccharide, or suppressive M2 states with interleukin-4. NFAT5, PPARG, and MECP2 were all upregulated during M2 polarization, and CRISPR-Cas9 knockdown of any of the three significantly reduced the M2 population, as measured by diminished expression of markers such as CD206 and CD163, while leaving M1 differentiation largely intact. The same results held when macrophages were exposed to conditioned medium from colorectal cancer cells, a more tumor-like stimulus, reinforcing the link between these factors and pro-tumorigenic polarization.

The most conceptually significant part of the study concerned how the relevant enhancers actually work. Enhancers can activate genes in two ways: by opening promoter chromatin so that the transcription preinitiation complex can assemble, or by releasing RNA polymerase II that has initiated transcription but stalled just downstream of the transcription start site. Classifying TAM-associated genes by whether their expression tracked promoter or enhancer accessibility, the researchers found that 71.2 percent of genes linked to Macro-SPP1 regulatory elements were so-called type E genes, whose promoters remained constitutively accessible across all cell states while only their enhancers changed. These type E genes, which included CCL2, HIF1A, CD274, and NFAT5, were also significantly enriched for fine-mapped cancer-risk variants from genome-wide association studies, unlike regulatory elements linked to other macrophage populations.

Constitutively open promoters with dynamically regulated enhancers are precisely the configuration in which pause-release operates, a paradigm well established in embryonic development. Re-analysis of RNA polymerase II ChIP-seq data from mouse macrophages confirmed the prediction: type E genes showed pronounced polymerase pausing in resting macrophages, and that pausing dropped sharply upon interleukin-4 stimulation, the signal that drives alternative, tumor-supportive polarization. The team then dissected the gene HIPK2, a type E gene whose product dampens NF-κB signaling. Disrupting individual HIPK2 enhancers with CRISPR-Cas9 reduced HIPK2 expression, lowered M2 marker expression at both RNA and protein levels, and, critically, increased polymerase occupancy at the transcription start site while depleting it from the gene body, the molecular fingerprint of enhanced pausing. Disrupting an upstream enhancer of NFAT5 produced a parallel loss of M2 polarization.

Taken together, the study sketches a sequential regulatory cascade for the monocyte-derived TAM route: AP-1 factors respond first to tumor infiltration, NF-κB drives an inflammatory phase, and that inflammation is then resolved into a pro-repair, immunosuppressive phenotype through PPARG and NFAT5 acting at super-enhancers that liberate paused polymerase at genes governing wound healing, angiogenesis, and immune suppression. The authors propose that blocking this pause-release mechanism with epigenetic or transcriptional inhibitors could strip tumors of their macrophage allies and potentially overcome resistance to checkpoint immunotherapy, a strategy conceptually distinct from simply depleting macrophages or blocking their recruitment.

The work carries caveats that the authors themselves acknowledge. The polymerase pausing data derive from mouse macrophages rather than human TAMs, so direct ChIP-seq in patient samples will be needed for definitive validation, and the survival analysis relies on bulk tumor data that cannot unambiguously assign gene expression to specific cell types. The in vitro polarization systems, while informative, are simplified stand-ins for the complex tumor microenvironment. Even so, by unifying single-cell chromatin maps across cancers, trajectory modeling, GWAS variant enrichment, and mechanistic perturbation experiments, the study delivers one of the most complete pictures to date of how human tumors reprogram macrophages at the level of cis-regulatory DNA, and it nominates a concrete set of enhancers, transcription factors, and polymerase dynamics as targets for the next generation of macrophage-directed cancer therapies.

Subject of Research: Epigenetic regulation of tumor-associated macrophage differentiation via enhancer-controlled RNA polymerase II pausing

Article Title: Pan-cancer epigenetic landscape of human tumor-associated macrophages reveals crucial enhancers governing their heterogenous formation by Pol II pausing modulation

Article References: Xiong, C., Ning, H., Wang, X., Xie, Y., Liu, Z., Hu, D., & Gao, X. (2026). Pan-cancer epigenetic landscape of human tumor-associated macrophages reveals crucial enhancers governing their heterogenous formation by Pol II pausing modulation. Journal of Advanced Research, 88, 929-945. https://doi.org/10.1016/j.jare.2026.01.048

Image Credits: AI Generated

DOI: 10.1016/j.jare.2026.01.048

Keywords: tumor-associated macrophages, enhancers, RNA polymerase II pausing, single-cell ATAC-seq, super-enhancers, NFAT5, PPARG, tumor microenvironment, macrophage polarization, cancer immunotherapy, epigenetics, pan-cancer analysis

The name MANTRA stands for Multi-view ANalysis with Tensor and matRix Alignment, and the framework was developed by Kevin De Azevedo, Yusuf Berk Oruc, and Florian Buettner, who contributed to the German Cancer Consortium and the Frankfurt Cancer Institute. Conceptually, MANTRA can be understood as a generalization of group factor analysis to tensor-valued data, or conversely, as an extension of classical tensor decomposition into the multi-view setting. It takes as input a collection of datasets, some three-dimensional and some two-dimensional, and decomposes them jointly into shared low-dimensional latent factors. Each mode of the data, whether patients, drugs, cytokines, cell types, or genes, receives its own embedding matrix, and products of these matrices yield interpretable loadings that expose interactions between the different dimensions.

What makes MANTRA technically distinctive is its Bayesian formulation with structured sparsity priors. The model places horseshoe priors on feature embeddings, a multi-scale shrinkage approach well suited to variational inference. Within this scheme, an Automatic Relevance Determination prior allows individual factors to be active in only a subset of data views, while local regularization encourages sparsity within each factor itself. This design has two practical consequences. First, unneeded factors are automatically driven toward zero, making the model remarkably robust when the true rank of the data is unknown, a common and frustrating problem in real applications. Second, the sparse factors highlight only the most relevant features, making the results biologically interpretable rather than opaque. Missing values, a perennial headache in clinical studies where not every omics layer can be profiled for every patient, are handled natively and in a principled manner.

The team validated MANTRA extensively on synthetic data before turning to real biology. In systematic benchmarks against PARAFAC, a state-of-the-art tensor decomposition implementation from the TensorLy package, MANTRA performed comparably in the high signal-to-noise regime and substantially better in the noisy, sparse, low-sample conditions that mimic real-world datasets. When the models were deliberately fitted with more factors than the true rank, MANTRA’s sparsity priors switched off the excess factors, whereas PARAFAC degraded. Scalability analyses on GPU hardware showed favorable scaling across sample counts, feature dimensions, and ranks, and sensitivity analyses confirmed that the choice of prior hyperparameters for the noise model leaves reconstruction accuracy stable across a broad range of data conditions.

The first real-world application focused on Chronic Lymphocytic Leukemia, a disease in which inter-patient heterogeneity strongly influences treatment outcomes. The researchers analyzed a published dataset of 192 primary CLL samples in which cell viability had been measured in response to 12 drugs under 17 different microenvironmental stimuli, complemented by bulk RNA sequencing of the same patients. This produced exactly the kind of mixed-order collection MANTRA was designed for: a third-order drug viability tensor of patients, drugs, and cytokines, alongside a second-order RNA-seq matrix of patients and genes. When trained on the drug tensor alone, MANTRA’s patient embeddings showed only a partial association with IGHV mutation status, one of the most important prognostic markers in CLL, distinguishing patients with typically aggressive unmutated disease from those with more indolent mutated disease.

The picture changed dramatically when the transcriptomic view was added. In the joint multi-view model, the patient embeddings separated sharply by IGHV status, and the drug loadings recovered known biological relationships, with drugs targeting the B-cell receptor, MAPK, and DNA damage response pathways clustering together. Quantitatively, the researchers used Leiden clustering of the learned latent space and measured alignment with clinical labels using adjusted Rand index and normalized mutual information, following standard practice from the single-cell integration literature. MANTRA outperformed MOFA+ and MOFA-FLEX, two leading matrix-based multi-omics tools, which had to flatten the three-dimensional drug tensor into stacks of two-dimensional matrices. MANTRA also separated patients by Trisomy 12, an intermediate-risk cytogenetic abnormality present in roughly ten to twenty percent of CLL patients, more accurately than the baselines, demonstrating that the model captures clinically meaningful axes of variation relevant to stratification and treatment response.

The second application pushed MANTRA into the single-cell arena, where the stakes were even higher. The team analyzed a multi-omics dataset from 18 pediatric patients with KMT2A-rearranged Acute Lymphoblastic Leukemia and 5 healthy donors, constructing two third-order tensors of patients, cell types, and features: one for single-cell RNA sequencing and one for single-cell ATAC sequencing. Because MANTRA tolerates missing values, it could use all 18 patients, whereas the specialized tool scITD, which requires every cell type to be observed in every donor, had to discard 7 of them. Both methods separated healthy donors from leukemia patients, but only MANTRA went further, splitting the patients into two distinct subgroups driven by a factor that was predominantly active in plasmacytoid dendritic cells, a rare immune cell population that turned out to be the surprising driver of disease heterogeneity.

Pathway analysis revealed what this plasmacytoid dendritic cell factor meant biologically. Genes loading strongly on the factor were enriched for interferon signaling and other immune-active pathways as well as endoplasmic reticulum stress programs, while lymphoid transcription factors such as ETV6 showed low loadings and genes typically implicated in T-cell lineage leukemia, including BCL2, MYC, and LYL1, showed high loadings. One patient subgroup carried a high-immune-activity, low-ER-stress program in their plasmacytoid dendritic cells, while the other occupied the opposite end of the spectrum. Critically, neither MOFA+, MOFA-FLEX, nor scITD detected this cell-type-specific stratification; all of their factors were dominated by blast cells. The finding was validated in an independent cohort of 7 pediatric B-ALL patients and 4 healthy donors, and when MANTRA was run on pediatric T-ALL data as a negative control, no such factor appeared, confirming that the plasmacytoid dendritic cell program is specific to B-cell acute lymphoblastic leukemia.

The linear structure of MANTRA is both its strength and its acknowledged limitation. Linear factor models are inherently interpretable, allowing biologists to trace each factor back to specific cell types, drugs, and pathways through the generalized loadings matrix. But capturing non-linear relationships would require kernel methods or deep architectures that would sacrifice precisely this transparency. The authors also note that in datasets where structural dependencies between dimensions are negligible, or where extreme sparsity and noise prevail, simpler matrix factorizations may remain equally effective or more stable. Identifiability and rotational invariance, perennial issues for all latent variable models, also apply here. Still, the researchers point toward supervised extensions of the framework as a promising future direction, albeit one that brings new overfitting challenges.

The implications reach well beyond leukemia. Drug screens that measure patient-derived cells across drugs and microenvironmental conditions, single-cell atlases that profile multiple tissues per donor, and multi-modal studies combining expression with chromatin accessibility all generate data with exactly the mixed-order tensor structure that MANTRA handles natively. By refusing to flatten higher-order structure into matrices, and by handling missing data through principled Bayesian inference rather than ad hoc imputation, the tool opens a window onto biological variation that existing methods systematically miss. The code is publicly available on GitHub, implemented in Python using the probabilistic programming library Pyro, and the discovery that a rare immune cell type defines clinically relevant leukemia subgroups offers a vivid demonstration of what becomes visible when algorithms finally learn to see data in all of its dimensions.

Subject of Research: Interpretable Bayesian integration of mixed-order multi-omics tensors for leukemia patient stratification

Article Title: Interpretable multi-omics integration across mixed-order tensors with MANTRA

Article References: De Azevedo, K., Oruc, Y. B., & Buettner, F. (2026). Interpretable multi-omics integration across mixed-order tensors with MANTRA. Molecular Systems Biology, 22(9), 1415-1429. https://doi.org/10.1038/s44320-026-00223-8

Image Credits: AI Generated

DOI: 10.1038/s44320-026-00223-8

Keywords: MANTRA, tensor decomposition, multi-omics, Bayesian inference, chronic lymphocytic leukemia, acute lymphoblastic leukemia, drug response, single-cell RNA-seq, single-cell ATAC-seq, plasmacytoid dendritic cells, structured sparsity, patient stratification

The work, led by Mitali U. Sheth, Wen-Lyong Qiu, Xiang Ru Ma and colleagues, addresses a gap that has widened as single-cell genomics has exploded in scale. Technologies such as single-cell RNA sequencing can now measure gene expression in hundreds of thousands of individual cells simultaneously, and single-cell ATAC-seq can profile chromatin accessibility, revealing which regulatory regions are open and potentially active in each cell. But these measurements, taken separately, do not by themselves reveal regulatory wiring. An enhancer may be accessible in a neuron, yet without evidence linking its activity to a target gene’s expression, its functional role remains speculative. Traditional approaches to establishing enhancer–gene links, such as CRISPR-based enhancer deletion or perturbation screens, are powerful but labor-intensive and difficult to scale across the enormous diversity of cell types in complex tissues.

The framework developed by the team leverages a key statistical insight: regulatory interactions leave fingerprints in the natural variation of gene expression and chromatin state across cells. If an enhancer genuinely controls a gene, then across the thousands of cells captured in a single-cell dataset, variation in that enhancer’s activity should be correlated with variation in the target gene’s expression, with effects that appear at the expected genomic distance and that are shaped by transcription factor binding motifs embedded in the enhancer sequence. By aggregating these weak signals across large cell populations and modeling them jointly, the method can distinguish true regulatory links from the many spurious correlations that arise from shared cellular states, cell cycle effects, or batch artifacts — confounders that have historically plagued correlation-based enhancer–gene inference.

A critical component of the approach is its handling of confounding at the level of cell identity. Because all genes in a given cell type tend to rise and fall together, naive correlation analysis assigns enhancers promiscuously to nearby genes, inflating the apparent regulatory network. The new method explicitly models and removes this shared component of variation, isolating the residual, enhancer-specific signal that reflects direct regulation. It also incorporates prior biological knowledge, including enhancer–promoter distance constraints, chromatin contact frequency data from technologies such as Hi-C and promoter capture assays, and sequence features that indicate which transcription factors bind a given enhancer. The result is a ranked set of candidate enhancer–gene pairs for each cell type, each accompanied by a quantitative confidence score that researchers can use to prioritize downstream experimental validation.

To benchmark the framework, the authors compared its predictions against gold-standard perturbation data — instances in which individual enhancers had been experimentally deleted or repressed and the resulting change in target gene expression measured directly. Across the validation sets, the computational predictions recovered a substantial fraction of experimentally confirmed enhancer–gene pairs while maintaining specificity, meaning they did not drown the true positives in a sea of false links. Notably, the method correctly identified many cases in which the nearest gene is not the actual target, a phenomenon that single-cell perturbation screens have increasingly shown to be common. Enhancers frequently skip over adjacent genes to contact promoters further away, and distance-based or nearest-gene assumptions systematically miss these long-range relationships.

One of the most striking demonstrations of the framework’s power comes from its application to disease genetics. When the researchers overlaid genome-wide association study summary statistics for a range of traits and diseases onto their enhancer–gene maps, they were able to trace risk variants to plausible target genes in a cell-type-specific manner. This step is where the method’s value becomes tangible for translational research. A noncoding variant associated with, say, an autoimmune disorder may lie within an enhancer active only in a specific subset of immune cells; by linking that enhancer to its gene targets computationally, researchers gain immediate hypotheses about the molecular mechanism of disease risk and the cellular context in which it operates. The maps effectively convert a list of statistically associated genomic coordinates into a functional annotation, complete with directionality and tissue relevance.

The study also illustrates how the same regulatory architecture can produce different outcomes in different cell types. The authors found that a substantial proportion of enhancers exhibit cell-type-specific target preferences: a given regulatory element may drive expression of one gene in a cortical neuron and an entirely different gene in an astrocyte, even when both cell types share the same underlying genome. This context dependence has major implications for interpreting both normal development and disease. It means that a single variant in a pleiotropic enhancer can contribute to multiple phenotypes through distinct target genes depending on where it is active, and it underscores why bulk-tissue studies, which average signals across mixed cell populations, often fail to resolve the relevant biology.

Technically, the framework operates on paired or unpaired single-cell multi-ome data, in which gene expression and chromatin accessibility are profiled either in the same individual cells or in parallel samples from the same biological condition. The method constructs a graph-based representation of cell-to-cell similarity, smooths sparse single-cell measurements over this graph to recover signal lost to dropout and shallow sequencing depth, and then evaluates enhancer–gene relationships using a regularized regression model that jointly considers all candidate enhancers within a defined genomic window around each gene. Regularization is essential: with tens of thousands of candidate enhancers competing to explain each gene’s expression, the model must penalize complexity to avoid overfitting noise. The learned weights, combined with the motif and chromatin contact priors, yield the final interaction scores. The authors have made the pipeline available to the community, and its computational design allows it to scale to the atlas-sized datasets — millions of cells spanning dozens of tissues — that are now being generated by consortia worldwide.

The implications extend well beyond human genetics. Enhancer–gene maps of this kind provide a substrate for studying gene regulatory network evolution, allowing comparative analyses of how regulatory wiring differs between species or between healthy and diseased states. In cancer genomics, where structural variants frequently rewire enhancer–promoter contacts to activate oncogenes, cell-type-resolved regulatory maps could help identify which tumors depend on which enhancer hijacking events. In developmental biology, the framework offers a way to trace how transcriptional programs are controlled as cells progress through differentiation, capturing transient regulatory relationships that exist only in rare intermediate cell states — populations too small and too fleeting to study with bulk methods.

Limitations remain, and the authors are careful to acknowledge them. Correlation-based inference, however sophisticated, cannot fully substitute for direct perturbation; some predicted interactions will prove false when tested experimentally, and some true interactions may be missed if the relevant enhancing activity is rare or condition-specific. Single-cell datasets also under-sample rare cell types, meaning that regulatory maps for those populations will be less complete. Moreover, chromatin accessibility is an imperfect proxy for enhancer activity: an open region is not necessarily an active enhancer, and methods that read out actual enhancer transcription, such as single-cell eRNA detection, could add further resolution in the future. The integration of the computational maps with systematic CRISPR perturbation screens — using the predictions to guide which enhancers to test — represents a promising hybrid strategy that combines scale with causal certainty.

What the study ultimately delivers is a change in the default assumption researchers can make when confronted with a noncoding variant. Where the fallback was once the nearest gene, it can now be a cell-type-specific, confidence-scored list of candidate targets derived from the collective behavior of thousands of single cells. As single-cell atlases continue to grow and as perturbation technologies mature, frameworks like this one are positioned to become standard infrastructure for interpreting the regulatory genome — the layer of DNA that, despite encoding no proteins, orchestrates when, where, and how much every gene is expressed. In bridging the gap between variant catalogs and mechanism, the work brings the field measurably closer to the long-promised era of actionable noncoding genetics.