The human gut is a complex barrier organ that supports essential physiological processes, including digestion, water and nutrient absorption, as well as waster elimination. Inflammatory bowel disease (IBD) is a group of prevalent chronic disorders that disrupt the normal structures and functions of the gut. Without effective management, IBD greatly compromises a patient's quality of life. Studies addressing the current challenges in understanding IBD are therefore critical to improve the clinical management. Healthy intestinal mucosa collected during routine endoscopic screening, can serve as a valuable control to support disease-focused studies.
Datasets (1)
Healthy intestinal mucosa: a control reference for inflammatory bowel disease103,385 cells
The initial wave of the COVID-19 pandemic provided a unique opportunity to study, at single-cell resolution, how near-universal exposure to the same pathogen can lead to either effective or dysfunctional immune responses in humans. Although single-cell RNA-sequencing studies have characterized features of COVID-19 cellular immunity, they have not shown how treatment with immunomodulators changes these features at single-cell resolution, or which features persist into convalescence. In this study, we analyzed 2.5 million circulating immune cells from 428 patients across time points (840 PBMC samples), encompassing three contemporaneous SARS-CoV-2 cohorts: acutely infected patients across five WHO disease severity levels and three time points, patients from the first randomized control trial to study the efficacy of tocilizumab in the management of COVID-19, and convalescent patients three months after infection. We used linear modeling to integrate multiple data types – including single-cell RNA-seq, CITE-seq, TCR and BCR sequencing, viral load measurements, viral neutralization assays, detection of 64 autoantibodies and 7 antiviral antibodies, HLA genotype data, and serum proteomics covering 1,463 targets – to derive the most comprehensive view to-date of the biological features of COVID-19 disease severity. Our findings demonstrate that myeloid-derived suppressor cells (MDSCs) act as a key immunologic pivot point in severe COVID-19. Myeloid dysfunction, which is marked by impaired antigen presentation, drives a non-productive adaptive immune response, as reflected by reduced expression of B and T cell gene programs involved in antigen recognition, immune synapse formation, and cytotoxicity. Severe disease is also linked to autoantibodies targeting type I interferons, influenced by specific HLA-DQB1 allelic variants, and strongly correlated with serum IL-6 levels. Tocilizumab treatment appears to eliminate CLU-expressing MDSCs and ISG-positive myeloid subsets, restore antigen presentation, and reactivate productive adaptive immunity. These changes align with improved clinical outcomes and laboratory measures, including reduced C-reactive protein. While many immunologic abnormalities in acute severe COVID-19 resolve during convalescence 3-months post-infection, we observed persistently high ICOS expression in regulatory T cells, potentially linking acute infection to chronic post-COVID syndromes. Overall, we define distinct innate and adaptive host immune responses associated with acute, IL-6–responsive, and convalescent SARS-CoV-2 infection. Our multimodal and high-dimensional dataset with curated clinical metadata provides a foundational and clinically relevant resource for modeling host immune response biology in health and disease. This CAP publication contains the single-cell RNA-seq data and cell annotations for the acute cohort.
Datasets (1)
Acute COVID-19 cohort - peripheral blood mononuclear cells across disease severity and time2,178,129 cells
The Human Cell Atlas Kidney (HCKA) v1 is a comprehensive single-cell reference of the human kidney, integrating ~1 million cells from 18 datasets with healthy tissue and multiple disease states, including AKI, glomerulonephritis, transplant rejection, polycystic kidney disease, and diabetic kidney disease. Using a harmonized analysis framework, the atlas defines 123 kidney cell populations. This resource provides a high-resolution map of kidney cellular diversity and disease mechanisms, serving as a reference for the study of kidney function and pathology. The atlas includes both single-cell and single-nucleus datasets, which are harmonized at the cell type annotation level to align identities across modalities while preserving modality-specific data structures, enabling unbiased cross-platform comparisons.
Datasets (2)
Single cell human kidney cell atlas811,098 cells
Single nucleus human kidney cell atlas215,308 cells
The Human Gut Cell Atlas (HGCA) brings together the collective expertise of gut biologists, clinicians, and data generators to create a community-driven, healthy reference for the human gut. It integrates 27 datasets representing several sample collection and library preparation methods from donors across age groups, with samples dissected or biopsied from the duodenum to the rectum.
The gut is a complex organ, stratified along its longitudinal and radial axes and responsible for everything from nutrient absorption and excretion to immune regulation. The HGCA integrates 944,502 cells of 503 samples from 266 donors and provides community-informed annotations for a preliminary 105 cell types, including location-specific transcriptional states and rare intestinal populations.
The atlas will continue to be versioned to reflect expanding cell type, technological, and population diversity, supporting a high-quality, community-maintained resource for the field. We invite gut and single-cell researchers to put their critical thinking caps on and help shape the HGCA's annotations by voting, reasoning, and contributing directly here on CAP.
Datasets (5)
Human Gut Cell Atlas (HGCA) v1 - Myeloid50,047 cells
Human Gut Cell Atlas (HGCA) v1 - Stroma52,016 cells
Human Gut Cell Atlas (HGCA) v1 - Epithelial269,927 cells
Human Gut Cell Atlas (HGCA) v1 - Lymphoid572,400 cells
Human Gut Cell Atlas (HGCA) v1 - All Cells944,390 cells
The cornea, limbus, and sclera form the ocular surface, a critical interface that protects the eye, maintains visual function, and provides structural support. The cornea serves as the primary refractive surface and barrier against external pathogens. The limbus, located at the corneal-scleral junction, houses limbal stem cells essential for the regeneration of the corneal epithelium. The sclera provides structural integrity, protects the eye, and serves as the attachment site for extraocular muscles. Together, these three components play integral roles in sustaining ocular function and homeostasis. To capture the major cell types and subclasses of the ocular surface, we generated large-scale datasets using scRNA-seq and snRNA-seq technologies from healthy human donors spanning a wide range of ages, genders, and ethnicities. Additionally, publicly available datasets were integrated to expand the scope of the atlas. In total, the atlas includes over 1 million cells and nuclei from 102 donors. We identified 10 major cell classes and 31 types in the atlas, including rare populations such as limbal progenitor cells, Schlemn’s canel endothelium, and specific immune subtypes. The atlas also captures the full trajectory of epithelial differentiation. To further enhance our understanding of gene regulation, we generated a snATAC-seq atlas, profiling over 362k nuclei from 26 donors, including 8 developmental samples, facilitating the characterization of cell type-specific chromatin accessibility landscapes and regulatory elements. All datasets underwent a unified preprocessing pipeline and data integration to ensure consistency and quality. The atlas is publicly accessible through CELLxGENE, enabling interactive exploration of gene expression and regulatory landscapes. As part of the Human Cell Atlas initiative, this comprehensive ocular surface atlas marks a significant advance in understanding ocular tissue physiology and provides a valuable resource for studying ocular surface development, aging, and diseases such as corneal dystrophies, limbal stem cell deficiency, and scleral remodeling.
Datasets (4)
scRNA-seq of human ocular surface: cornea, limbus, sclera – fetal data included815,769 cells
scRNA-seq of human ocular surface: cornea, limbus, sclera - all cells577,857 cells
scRNA-seq of human ocular surface: cornea, limbus, sclera - epithelial cells250,321 cells
snRNA-seq of human ocular surface: cornea, limbus, sclera147,484 cells
Here we use single-cell RNA sequencing to compile a human breast cell atlas assembled from 55 donors that had undergone reduction mammoplasties or risk reduction mastectomies. From more than 800,000 cells we identified 41 cell subclusters across the epithelial, immune and stromal compartments. The contribution of these different clusters varied according to the natural history of the tissue. Age, parity and germline mutations, known to modulate the risk of developing breast cancer, affected the homeostatic cellular state of the breast in different ways. We found that immune cells from BRCA1 or BRCA2 carriers had a distinct gene expression signature indicative of potential immune exhaustion, which was validated by immunohistochemistry. This suggests that immune-escape mechanisms could manifest in non-cancerous tissues very early during tumor initiation. This atlas is a rich resource that can be used to inform novel approaches for early detection and prevention of breast cancer.
The survival and proper function of the neural retina depend on the support of two underlying tissues: the retinal pigment epithelium (RPE) and the choroid. Dysfunction in the RPE and choroid can lead to a range of inherited and complex diseases. To better characterize the cellular, transcriptomic, and epigenomic heterogeneity and dynamics within the RPE and choroid, we conducted single-cell and single-nucleus RNA sequencing (sc/snRNA-seq) as well as single-nucleus ATAC sequencing (snATAC-seq) on donor tissue samples. By integrating newly generated and previously published datasets, we developed version 1 of the human RPE and choroid cell atlas. This atlas includes scRNA-seq data from 205,925 cells from 67 donors and snRNA-seq data from 742,625 nuclei from 54 donors. Additionally, we incorporated snRNA-seq data from 136,555 nuclei from developmental RPE/choroid tissues spanning post-conception weeks (PCW) 10-23. This comprehensive dataset enabled the identification of 15 major cell classes, which were further subdivided into 25 cell types. Additionally, open chromatin profiling was performed on over 234,007 nuclei using snATAC-seq, providing an in-depth characterization of chromatin accessibility for each cell type. As part of the Human Cell Atlas project, this resource lays a robust foundation for advancing research into the biology of the RPE and choroid and the diseases associated with these tissues.
Datasets (9)
snRNA-seq of human retinal pigment epithelium and the choroid - v1715,619 cells
snRNA-seq of human retinal pigment epithelium and the choroid - all cells566,546 cells
scRNA-seq of human retinal pigment epithelium and the choroid182,151 cells
snRNA-seq of human retinal pigment epithelium and the choroid - stromal cells212,984 cells
snRNA-seq of human retinal pigment epithelium and the choroid - endothelial cells99,457 cells
snRNA-seq of human retinal pigment epithelium and the choroid - retinal pigment epithelial cells69,244 cells
snRNA-seq of human of the human retinal pigment epithelium and the choroid - Immune cells101,471 cells
snRNA-seq of human of the human retinal pigment epithelium and the choroid - Melanocytes61,751 cells
snRNA-seq of human of the human retinal pigment epithelium and the choroid - Schwann cells21,625 cells
Under normal conditions, the most significant expansion and differentiation of the adult mammary gland occurs in response to systemic reproductive hormones during pregnancy and lactation to enable milk synthesis and secretion to sustain the offspring. However, human mammary tissue remodelling that takes place during pregnancy and lactation remains poorly understood due to the challenge of acquiring samples. We report here single-cell transcriptomic analysis of 110,744 viable breast cells isolated from human milk or non-lactating breast tissue, isolated from nine and seven donors, respectively. We found that human milk largely contains epithelial cells belonging to the luminal lineage and a repertoire of immune cells. Further transcriptomic analysis of the milk cells identified two distinct secretory cell types that shared similarities with luminal progenitors, but no populations comparable to hormone-responsive cells. Taken together, our data offers a reference map and a window into the cellular dynamics that occur during human lactation and may provide further insights on the interplay between pregnancy, lactation and breast cancer.
We present the Human Liver Cell Atlas (HLiCA), integrating non-disease liver data from eight datasets across six research centers, covering 520,000+ cells from 110 donors. Developed in collaboration with the Human Cell Atlas Liver Bionetwork, the atlas incorporates expert-reviewed cell annotations, extensive feedback, and focused jamborees for refining myeloid and hepatocyte classifications.
Datasets (7)
Human liver cell atlas - all cells524,699 cells
Human liver cell atlas - hepatocyte cells228,368 cells
Human liver cell atlas - myeloid cells65,575 cells
Human liver cell atlas - lymphocyte cells103,347 cells
Human liver cell atlas - endothelia cells86,530 cells
Human liver cell atlas - mesenchyme cells31,278 cells
Human liver cell atlas - cholangiocyte cells8,811 cells
The breast is a dynamic organ whose response to physiological and pathophysiological conditions alters its disease susceptibility, yet the specific effects of these clinical variables on cell state remain poorly annotated. We present a unified, high-resolution breast atlas by integrating single-cell RNA-seq, mass cytometry, and cyclic immunofluorescence, encompassing a myriad of states. We define cell subtypes within the alveolar, hormone-sensing, and basal epithelial lineages, delineating associations of several subtypes with cancer risk factors, including age, parity, and BRCA2 germline mutation. Of particular interest is a subset of alveolar cells termed basal-luminal (BL) cells, which exhibit poor transcriptional lineage fidelity, accumulate with age, and carry a gene signature associated with basal-like breast cancer. We further utilize a medium-depletion approach to identify molecular factors regulating cell-subtype proportion in organoids. Together, these data are a rich resource to elucidate diverse mammary cell states.
Datasets (1)
Mapping hormone-regulated cell-cell interaction networks in the human breast at single-cell resolution (Murrow et al. 2022)86,136 cells
The integrated Human Breast Cell Atlas (iHBCA) v1.0 refines iHBCA introduced by Reed et al. 2024 (Nature Genetics) by harmonizing 7 published single-cell breast studies (Gray 2022, Kumar 2023, Murrow 2022, Nee 2023, Pal 2021, Reed 2024, Twigger 2022) into a cross-study reference of 2,128,505 cells from 287 donors. Cells are integrated with scVI and manually annotated at a broad L1 resolution (11 cell type groups spanning epithelial, stromal, immune, and endothelial compartments). Donors span reproductive history (nulliparous, parous, lactating), age ranges from young adult through late adult, and germline BRCA1/BRCA2 mutation carriers alongside non-carrier controls.
Datasets (1)
Integrated Human Breast Cell Atlas v1.0 - all cells2,128,505 cells
Tendons are transitional tissues linking muscle to bone, enabling locomotion and fine motor control. The cellular biology across the Achilles tendon unit is poorly understood, yet is critical for interpreting normal function and pathological changes across its microanatomically defined functional zones. We generated a spatially resolved transcriptomic atlas of adult (age 45–76) nontendinopathic human Achilles tendon, sampling the tendon-bone junction (enthesis), midbody, myotendinous junction, and adjoining muscle. Six fibroblast subtypes were identified, with distinct transcriptional profiles and spatial distributions, suggesting specialized functional roles across the tendon. Two dominant fibroblast types were specifically positioned in the tendon midsubstance and paratenon (vessel-rich region surrounding the tendon fibrils); other populations included perineural, myotendinous junction-specific, muscle-specific, and lining-layer fibroblasts. These findings demonstrate how cellular diversity across a transitional tissue may underlie microanatomical-specific roles. This atlas provides a foundation for understanding cellular functions across the tendon and adjoining muscle and will be essential for comparisons with diseased tissue, identifying pathogenic mediators and treatment targets for autoimmune and degenerative pathologies of the Achilles tendon.
Allogeneic hematopoietic stem cell transplantation (HSCT) is a life-saving treatment for hematologic malignancies, but long-term survivors present with lower muscle mass and functional capacity. In adult HSCT survivors 10-20 years after treatment, single nucleus RNA sequencing uncovered elevated XRRA1 expression levels in all muscle nuclei populations, which was retained in primary muscle stem cell cultures. HSCT survivors were characterized in vivo by impaired neuromuscular innervation that associated with muscle weakness, and lower muscle stem cell neurotrophic action. Despite these impairments, the molecular and physiological responses to heavy resistance training (HReT) were preserved in HSCT survivors, as demonstrated in a pre-registered clinical trial (ClinicalTrials.gov: NCT04922970). After 12 weeks of HReT, gains in muscle mass and strength were similar in HSCT survivors and healthy controls. In addition, we observed that ∼9% of muscle-resident immune cells persist into adulthood and that bone marrow derived cells do not adopt alternative cell fates in muscle tissue, resolving long-standing questions in human muscle biology. Together, these findings uncover molecular mechanisms of HSCT sequelae in muscle nuclei and muscle stem cells, which, importantly, can at least partly be overcome by mechanical loading. Given the growing population of HSCT survivors and the multitude of benefits of HReT for all organ systems, our findings support the importance of HReT in this population to promote healthspan.
Datasets (1)
Healthy human vastus lateralis snRNAseq at rest and 7 days post exercise180,419 cells
The myotendinous junction (MTJ) is a specialized domain of the multinucleated myofibre that is faced with the challenge of maintaining robust cell–matrix contact with the tendon under high mechanical stress and strain. Here, we profiled 24,124 nuclei in semitendinosus muscle–tendon samples from three healthy males by using single-nucleus RNA sequencing (snRNA-seq), alongside spatial transcriptomics, to gain insight into the genes characterizing this specialization in humans. We identified a cluster of MTJ myonuclei represented by 47 enriched transcripts, of which the presence of ABI3BP, ABLIM1, ADAMTSL1, BICD1, CPM, FHOD3, FRAS1 and FREM2 was confirmed at the MTJ at the protein level in immunofluorescence assays. Four distinct subclusters of MTJ myonuclei were apparent, comprising two COL22A1-expressing subclusters and two subclusters lacking COL22A1 expression but with differing fibre type profiles characterized by expression of either MYH7 or MYH1 and/or MYH2. Our findings reveal distinct myonuclei profiles of the human MTJ, which represents a weak link in the musculoskeletal system that is selectively affected in pathological conditions ranging from muscle strains to muscular dystrophies.
Mechanical loading drives structural and functional improvements in muscle and tendon, protecting against injury at their interface – the myotendinous junction (MTJ) – and within the tendon matrix. However, the early cellular and molecular events that initiate these adaptations in humans remain poorly understood. To investigate this, we applied single nucleus RNA sequencing and in situ hybridization to map the acute transcriptional response of the human muscle-tendon unit to a single bout of eccentric resistance exercise, with a focus on extracellular matrix (ECM) regulation. We identified four transcriptionally distinct fibroblast subtypes expressing key ECM components, including COL1A1 and DCN. Three of these subtypes were localized to tendon and responded to exercise: two were spatially restricted to the collagen fascicles or the MTJ, while the third, enriched in the interfascicular matrix (IFM), exhibited the strongest response. This IFM population, marked by PDGFRA, upregulated PRG4 and VCAN, ECM genes linked to tissue lubrication and resilience. In parallel, exercise induced dynamic ECM regulation in myonuclei, particularly in a distinct subset of type II myonuclei at the MTJ that expanded in number and robustly upregulated COL22A1, a collagen essential for MTJ integrity. Together, these findings uncover a spatially organized, cell type-specific program of ECM remodeling in response to mechanical load, offering new insight into the early molecular events of human muscle-tendon adaptation.
Datasets (1)
Human semi-tendinosus and gracilis with and without acute exercise nuclei atlas - hamstring tendon nuclei61,215 cells
Our atlas represents a collaborative international effort to successfully integrate and annotate multiple large scale published and unpublished PBMC datasets spanning healthy and COVID patients. Using rigorous metrics for quality control we were formed a final integrated dataset of 8.4M cells. Consensus annotation was performed across multiple sites at high resolution, and ultimately defined 192 individual immune cell states in the peripheral blood.
Datasets (11)
Human PBMC atlas - B memory cells337,659 cells
Human PBMC atlas - plasma cells37,117 cells
Human PBMC atlas - B naive cells463,552 cells
Human PBMC atlas - T CD8 naive cells540,294 cells
Human PBMC atlas - TCD4 effector and nonnaive cells749,849 cells
Human PBMC atlas - TCD4 naive cells1,625,425 cells
Human PBMC atlas - T CD8 effector and nonnaive cells and innate lymphoid cells including NK cells2,216,128 cells
Human PBMC atlas - T/NK proliferating cells102,483 cells
Adipose tissue (AT) is a highly dynamic organ serving metabolic, endocrine, and immune functions that collectively regulate systemic energy homeostasis and thermogenesis. It is organized into anatomically distinct depots — the most extensively studied being abdominal subcutaneous (ASAT) and visceral adipose tissue (AVAT) — but also dispersed across many other tissues and organs, such as cervical subcutaneous adipose tissue (CSAT). The functional importance of AT is underscored by the global burden of adipose-related pathophysiology. Excess adiposity, which defines obesity, now affects over one billion individuals worldwide and is a major risk factor for metabolic dysfunction and common comorbidities, including type 2 diabetes, cardiovascular disease, cancer, and premature mortality — imposing a substantial clinical and economic burden globally. Deciphering the cellular and molecular mechanisms governing human white adipose tissue (WAT) across health and disease, anatomical locations, sex, and age is therefore of critical importance.
Here we present a comprehensive, integrated single-nucleus transcriptomic atlas of human WAT comprising 1,037,815 nuclei from 184 donors and 411 samples, spanning abdominal (ASAT and AVAT) and cervical (CSAT) depots across 22 studies, of which 7 remain currently unpublished (AlZaim 2026, Emont 2022, Grothen 2026, Hinte 2024, Jalkanen 2026, Lazarescu 2025, Loft 2026, Massier 2023, Reinisch 2025, Sun 2020, Wang 2024). The HATA was developed in accordance with Human Cell Atlas standards, incorporating robustness to technical variation, systematic benchmarking of integration methods, and consensus cell type annotation by members of the HCA adipose tissue community. As a standardized reference with unified cell type nomenclature, the HATA enables high-resolution dissection of AT biology and provides a foundation for uncovering the molecular mechanisms underlying depot-specific function and pathological remodeling in obesity and related metabolic diseases.
Datasets (10)
Human Adipose Tissue Atlas v1.0 — All cells1,037,815 cells
Human Adipose Tissue Atlas v1.0 — Abdominal Subcutaneous WAT555,527 cells
Human Adipose Tissue Atlas v1.0 — Abdominal Visceral WAT433,708 cells
Human Adipose Tissue Atlas v1.0 — Adipocytes249,354 cells
Human Adipose Tissue Atlas v1.0 — ASPCs225,219 cells
Human Adipose Tissue Atlas v1.0 — Vascular cells210,838 cells
Human Adipose Tissue Atlas v1.0 — Myeloid cells181,693 cells
Human Adipose Tissue Atlas v1.0 — Mesothelial cells118,373 cells
Human Adipose Tissue Atlas v1.0 — Cervical Subcutaneous WAT48,580 cells
Human Adipose Tissue Atlas v1.0 — Lymphoid cells49,387 cells
While it is well established that the cellular composition of white adipose tissue (WAT) varies between depots, the functional relevance of this heterogeneity remains unclear. By combining spatial and single-nucleus RNA sequencing, we provide a comprehensive map of subcutaneous and visceral (omental, mesenteric, mesocolic, and epiploic) WAT in both men and women. Our analyses reveal shared features, such as the spatial organization of adipogenesis, alongside depot-specific characteristics, including distinct cell-type enrichments and unique cell-cell communication routes. Epiploic WAT stands out by harboring high proportions of serum amyloid A expressing fat cells (encoded by SAA1/SAA2) and several leukocyte populations. Through mechanistic studies, we demonstrate that adipocyte SAA1/SAA2 expression is induced by inflammatory signals, including lipopolysaccharide, and that SAA1 activates immune responses in adipose-resident myeloid cells. Collectively, our findings suggest that visceral WAT exhibits distinct cytoarchitectural properties, with those located near the colon adapting by developing specialized adipocytes and immune cell populations.
Datasets (1)
Subcutaneous, omental, mesenteric, mesocolic, and epiploic adipose tissue (Jalkanen et al., 2026, Cell Met)255,273 cells
Human white adipose tissue undergoes major remodelling during sustained weight gain that may compromise tissue function and drive cardiometabolic comorbidities. Although weight loss reverses many of these complications, the cellular and molecular adaptations of adipose tissue to different weight loss interventions are poorly understood. Here we profiled abdominal subcutaneous adipose tissue (SAT) from men and women living with severe obesity using single-nucleus RNA sequencing (snRNA-seq). Biopsies from the same individuals were collected at baseline, after modest lifestyle-induced (8–10%) weight loss, and again after bariatric surgery-induced (20–45%) weight loss. Lifestyle-induced weight loss activated proadipogenic gene programmes in progenitor cells, indicating early beneficial effects on SAT. Subsequent surgery-induced weight loss drove profound compositional and transcriptional remodelling of SAT, including increased vascularization and marked reduction of myeloid cell populations. Collectively, our study indicates that following major and sustained weight loss, SAT from individuals with severe obesity has the capacity to return to a state comparable to that observed in lean individuals.
In this study we characterized the diversity of cell types in human skeletal muscle (intercostal muscle) across the lifespan (8 young donors ~20-40 yrs, 9 aged donors ~60-75 yrs) using single-cell and single-nuclei seiquencing totalling ~ 183, 000 cells. We leveraged the aforementioned dataset to systematically study change in cell type composition and gene expression between young and old adults, highlighting changes in the major skeletal muscle compartments, including muscle satellite cells, myofiber and muscle microenvironment including stromal, immune and vascular cell types. We identified decreased ribosome biogenesis in TNFRSF12A+ MuSCs, CCL2-driven inflammatory remodelling of the aged microenvironment, expansion of NMJ-accessory nuclei as a denervation response, and a slow-to-fast myonuclear shift plus MYH8+ regenerative myocytes compensating fast-twitch loss. Additionally, we generated a complementary mouse muscle aging dataset by profiling hindlimb muscles from young (n = 5, 3 months) and old mice (n = 3, 19 months), using single-cell and single-nucleus sequencing for comparison. Resource: https://www.muscleageingcellatlas.org/
Datasets (1)
Human intercostal skeletal muscle aging atlas (scRNA-seq and snRNA-seq, young vs aged)183,161 cells
Training can improve insulin sensitivity in individuals with type 2 diabetes, but a clear understanding of the mechanisms remains elusive. To further our knowledge in this area, we aimed to examine the effect of type 2 diabetes and of high-intensity interval training (HIIT) on the nuclear transcriptional response in skeletal muscle. We performed single-nucleus RNA-sequencing (snRNA-seq) and immunofluorescence analysis on muscle biopsies from the trained and the untrained legs of participants with and without type 2 diabetes, after 2 weeks of one-legged HIIT on a cycle ergometer. Surprisingly, the type 2 diabetes condition only seemed to have a minor effect on transcriptional activity in myonuclei related to major metabolic pathways when comparing the untrained legs. However, while in particular the type IIA myonuclei in the control group displayed a considerable metabolic response to HIIT, with increases in genes related to glycogen breakdown and glycolysis primarily in the type IIA myonuclei of the trained leg, this response was blunted in the diabetes group, despite a marked increase in glucose clearance in both groups. Additionally, we observed that fibre type distribution assessed by immunofluorescence significantly correlated with the proportion of myonuclei in the snRNA-seq analysis. In conclusion, the type 2 diabetes condition blunts the metabolic transcriptional response to HIIT in the type IIA myonuclei without affecting the improvement in insulin sensitivity. Additionally, our results indicate that snRNA-seq can be used as a surrogate marker for fibre type distribution in sedentary middle-aged adults.
This dataset is a subset containing the healthy donors from the study.
Datasets (1)
snRNAseq on trained and untrained human skeletal muscle from healthy donors52,597 cells
As part of a larger study on understanding the relationship between genetic variants and their functional consequences, here we profile single nucleus gene expression in skeletal muscles from six healthy donors.
Datasets (1)
Single-nucleus transcriptomic profiling of human skeletal muscle from healthy adults40,769 cells
White adipose tissue, once regarded as morphologically and functionally bland, is now recognized to be dynamic, plastic and heterogenous, and is involved in a wide array of biological processes including energy homeostasis, glucose and lipid handling, blood pressure control and host defence1. High-fat feeding and other metabolic stressors cause marked changes in adipose morphology, physiology and cellular composition1, and alterations in adiposity are associated with insulin resistance, dyslipidemia and type 2 diabetes2. Here we provide detailed cellular atlases of human and mouse subcutaneous and visceral white fat at single-cell resolution across a range of body weight. We identify subpopulations of adipocytes, adipose stem and progenitor cells, vascular and immune cells and demonstrate commonalities and differences across species and dietary conditions. We link specific cell types to increased risk of metabolic disease and provide an initial blueprint for a comprehensive set of interactions between individual cell types in the adipose niche in leanness and obesity. These data comprise an extensive resource for the exploration of genes, traits and cell types in the function of white adipose tissue across species, depots and nutritional conditions.
Datasets (1)
Emont 2022 — abdominal subcutaneous and visceral adipose tissue137,684 cells
We present a single-nucleus RNA sequencing (snRNA-seq) dataset of human skeletal muscle derived from 16 donors in the Singapore Adult Metabolism Study (SAMS1&2) cohort. The samples were collected from vastus lateralis muscle biopsies of individuals across multiple Asian ancestries (Chinese, Malay, and Indian), encompassing both lean and obese metabolic states, as well as before and after weight-loss. This dataset captures the cellular heterogeneity of human skeletal muscle and provides a resource for investigating ancestry-specific and metabolic state-associated transcriptional variation. In particular, the inclusion of diverse Asian populations addresses a critical gap in current human muscle transcriptomic references, which are predominantly derived from European cohorts.
Additional details of the SAMS2 cohort can be found here: https://www.cell.com/cell-genomics/fulltext/S2666-979X(25)00207-1
Please see the external URLS for SAMS1 cohort related publications.
Datasets (1)
Single-nucleus RNA-seq of Asian Skeletal Muscle Reveals Ancestry-specific Regulatory Programs Across Metabolic States62,056 cells
Human cells share an identical genome yet exhibit diverse functions shaped by their tissue environment and cell-type-specific regulatory programs. While single-cell transcriptomic atlases have mapped cellular diversity across many human tissues, comparable atlases of regulatory elements at single-cell resolution remain scarce, limiting our understanding of how regulatory codes vary across organs. To address this gap, we generated high-quality 10x Genomics Multiome profiles from 21 human tissues collected from 4 GTEx donors, jointly measuring gene expression and chromatin accessibility in over 450,000 single cells. This cross-organ resource enables systematic characterization of transcription factor activity, cis-regulatory elements, and their coupling to gene expression across diverse cell types and tissue contexts. Together, these data illuminate the regulatory diversity of human single cells across organs and provide a foundational reference for studying tissue-specific gene regulation in health and disease.
Datasets (1)
Single-nucleus transcriptomic profiling of human skeletal muscle from healthy adults33,378 cells
To date, single-cell studies of human white adipose tissue (WAT) have been based on small cohort sizes and no cellular consensus nomenclature exists. Herein, we performed a comprehensive meta-analysis of publicly available and newly generated single-cell, single-nucleus, and spatial transcriptomic results from human subcutaneous, omental, and perivascular WAT. Our high-resolution map is built on data from ten studies and allowed us to robustly identify >60 subpopulations of adipocytes, fibroblast and adipogenic progenitors, vascular, and immune cells. Using these results, we deconvolved spatial and bulk transcriptomic data from nine additional cohorts to provide spatial and clinical dimensions to the map. This identified cell-cell interactions as well as relationships between specific cell subtypes and insulin resistance, dyslipidemia, adipocyte volume, and lipolysis upon long-term weight changes. Altogether, our meta-map provides a rich resource defining the cellular and microarchitectural landscape of human WAT and describes the associations between specific cell types and metabolic states.
Datasets (1)
Massier 2023 — abdominal subcutaneous and omental adipose tissue38,081 cells
Precision medicine is still not considered as a standard of care in obesity treatment, despite a large heterogeneity in the metabolic phenotype of individuals with obesity. One of the strongest factors influencing the variability in metabolic disease risk is adipose tissue (AT) dysfunction; however, there is little understanding of the link between distinct cell populations, cell-type-specific transcriptional programs, and disease severity. Here, we generated a comprehensive cellular map of subcutaneous and visceral AT of individuals with metabolically healthy and unhealthy obesity. By combining single-nucleus RNA-sequencing data with bulk transcriptomics and clinical parameters, we identified that mesothelial cells, adipocytes, and adipocyte-progenitor cells exhibit the strongest correlation with metabolic disease. Furthermore, we uncovered cell-specific transcriptional programs, such as the transitioning of mesothelial cells to a mesenchymal phenotype, that are involved in uncoupling obesity from metabolic disease. Together, these findings provide valuable insights by revealing biological drivers of clinical endpoints.
Reducing body weight to improve metabolic health and related comorbidities is a primary goal in treating obesity. However, maintaining weight loss is a considerable challenge, especially as the body seems to retain an obesogenic memory that defends against body weight changes. Overcoming this barrier for long-term treatment success is difficult because the molecular mechanisms underpinning this phenomenon remain largely unknown. Here, by using single-nucleus RNA sequencing, we show that both human and mouse adipose tissues retain cellular transcriptional changes after appreciable weight loss. Furthermore, we find persistent obesity-induced alterations in the epigenome of mouse adipocytes that negatively affect their function and response to metabolic stimuli. Mice carrying this obesogenic memory show accelerated rebound weight gain, and the epigenetic memory can explain future transcriptional deregulation in adipocytes in response to further high-fat diet feeding. In summary, our findings indicate the existence of an obesogenic memory, largely on the basis of stable epigenetic changes, in mouse adipocytes and probably other cell types. These changes seem to prime cells for pathological responses in an obesogenic environment, contributing to the problematic ‘yo-yo’ effect often seen with dieting. Targeting these changes in the future could improve long-term weight management and health outcomes.
Human adipose depots are functionally distinct. Yet, recent single-nucleus RNA-sequencing (snRNA-seq) analyses largely uncovered overlapping/similar cell-type landscapes. We hypothesized that adipocytes subtypes, differentiation trajectories, and/or intercellular communication patterns could illuminate this depot similarity-difference gap. For this, we performed snRNA-seq of human subcutaneous/visceral adipose tissues (5/10 samples, respectively). Of 27,665 adipocyte nuclei in both depots, the majority were “classical”, namely- enriched in lipid metabolism pathways. However, we also observed “non-classical” adipocyte subtypes, enriched in immune-related, extracellular matrix deposition (fibrosis), vascularization/angiogenesis, or ribosomal processes. Pseudo-temporal analysis showed a developmental trajectory from adipose progenitor cells to classical adipocytes via non-classical adipocytes, suggesting that the classical state stems from loss, rather than gain, of specialized functions. Lastly, intercellular communication routes were consistent with the different inflammatory tone of the two depots. Jointly, these findings provide a high-resolution view into the contribution of cellular composition, differentiation, and intercellular communication patterns to human fat depot differences.
The mammalian brain consists of millions to billions of cells that are organized
into many cell types with specific spatial distribution patterns and structural and
functional properties1–3. Here we report a comprehensive and high-resolution
transcriptomic and spatial cell-type atlas for the whole adult mouse brain. The
cell-type atlas was created by combining a single-cell RNA-sequencing (scRNA-seq)
dataset of around 7 million cells profiled (approximately 4.0 million cells passing
quality control), and a spatial transcriptomic dataset of approximately 4.3 million
cells using multiplexed error-robust fluorescence in situ hybridization (MERFISH).
The atlas is hierarchically organized into 4 nested levels of classification: 34 classes,
338 subclasses, 1,201 supertypes and 5,322 clusters. We present an online platform,
Allen Brain Cell Atlas, to visualize the mouse whole-brain cell-type atlas along with the
single-cell RNA-sequencing and MERFISH datasets. We systematically analysed the
neuronal and non-neuronal cell types across the brain and identified a high degree of
correspondence between transcriptomic identity and spatial specificity for each cell
type. The results reveal unique features of cell-type organization in different brain
regions—in particular, a dichotomy between the dorsal and ventral parts of the brain.
The dorsal part contains relatively fewer yet highly divergent neuronal types, whereas
the ventral part contains more numerous neuronal types that are more closely related
to each other. Our study also uncovered extraordinary diversity and heterogeneity
in neurotransmitter and neuropeptide expression and co-expression patterns in
different cell types. Finally, we found that transcription factors are major determinants
of cell-type classification and identified a combinatorial transcription factor code that
defines cell types across all parts of the brain. The whole mouse brain transcriptomic
and spatial cell-type atlas establishes a benchmark reference atlas and a foundational
resource for integrative investigations of cellular and circuit function, development
and evolution of the mammalian brain.
Datasets (16)
Whole Mouse Brain Taxonomy, Near and corticothalamic projecting cortical glutamatergic neurons310,198 cells
Immunological processes that underpin human immune responses to therapeutics and vaccine components, such as vaccine adjuvants, remain poorly defined due to a paucity of models that faithfully recapitulate immune activation in lymphoid tissues. We describe precision-cut human lymph node (LN) slices as a functioning, architecturally preserved, full-organ cross-sectional model system. Using single-cell transcriptomics and multiplexed imaging, we explore early inflammatory response to a potent, clinically relevant liposomal vaccine adjuvant containing a TLR4-agonist and QS-21 saponin. Both TLR4 and NLRP3 inflammasome activation are involved in the direct initiation of the inflammatory response to adjuvant by monocytes and macrophages (Mon./Mac.) with secretion of interleukin (IL)-1β, but not IL-18, dependent on TLR4 signaling. Innate lymphoid cells, including natural killer cells, are indirectly activated by Mon./Mac.-produced cytokines, signaling downstream to B cells via interferon-γ secretion. Resident LN stromal populations, primed both directly and indirectly by vaccine adjuvant, are instrumental in mediating inflammatory cell recruitment, particularly neutrophils.
Datasets (2)
Human lymph node slices cultured with and without ex vivo stimulation by liposomal TLR4 and QS-21 adjuvant109,297 cells
Human lymph nodes with partial T and B cell depletion77,525 cells
The integrated Human Lung Cell Atlas (HLCA) represents the first large-scale, integrated single-cell reference atlas of the human lung. It consists of over 2 million cells from the respiratory tract of 486 individuals, and includes 49 different datasets. It is split into the HLCA core, and the extended or full HLCA. The HLCA core includes data of healthy lung tissue from 107 individuals, and includes manual cell type annotations based on consensus across 6 independent experts, as well as demographic, biological and technical metadata. The datasets in the HLCA core were integrated using scANVI. The HLCA core can be used as a reference to map new datasets onto using scArches. The full HLCA includes 35 further datasets that include donors with various lung diseases. These datasets were mapped onto the core with scArches, and include disease annotations as well as cell type annotations transferred from the HLCA core onto the mapped datasets. Note that while the the HLCA includes an integrated, batch-correctd low-dimensional embedding, the gene counts themselves were not batch-corrected. Both the HLCA core and the full HLCA can be explored below. Detailed information about all metadata in the objects, as well as further HLCA-related information , can be found on the HLCA landing page: https://github.com/LungCellAtlas/HLCA. Raw counts are available in the downloaded .h5ad at adata.raw.X, in the downloaded .rds at seurat_object@assays$RNA@counts, and via CELLxGENE Census.
Datasets (1)
An integrated cell atlas of the human lung in health and disease (core)584,944 cells
We provide a large-scale reference atlas of human hematopoiesis with balanced representation of CD34+ stem and progenitor cells together with terminally differentiated populations. Our reference map of 263,159 hematopoietic cells provides us with deep portraits of lineage transitions in both early and late stages of human hematopoiesis.
HSPC annotations within this atlas were carefully validated to maximize concordance with immunophenotypically pure populations established through decades of functional studies. In particular, our transcriptionally defined HSC state is highly concordant with immunophenotypically defined Lin-CD34+CD38-CD45RA-CD90+CD49f+ long-term HSCs.
Please visit the Github page for our R package to access functions for mapping query scRNA-seq profiles of normal or leukemic hematopoietic cells onto BoneMarrowMap. The entire process is optimized to take <10 min total for mapping ~100,000 single cells on a personal laptop.
See also our Blood Cancer Discovery publication (Zeng et al, 2025) for more details on the atlas and an example of its application to leukemia.
Datasets (1)
BoneMarrowMap: A Balanced Single-cell Reference Atlas of Human Hematopoiesis263,159 cells
The optic nerve is essential for transmitting visual information from the retina to the brain, where it is processed into images. Damage to the optic nerve can cause vision loss, leading to conditions such as glaucoma, optic neuritis, and optic neuropathy. To advance our understanding of the transcriptomic and epigenetic landscapes, as well as the dynamic cellular processes of the optic nerve, we conducted single-nucleus RNA sequencing (snRNA-seq) and single-nucleus ATAC sequencing (snATAC-seq) on the optic nerve head (ONH) and optic nerve (ON) tissues from over 80 healthy human donor eyes spanning diverse ages, genders, and ethnic backgrounds. By integrating newly generated and previously published snRNA-seq datasets, we constructed a comprehensive ONH and ON cell atlas encompassing transcriptomic profiles of 959,629 nuclei. Our analysis identified 25 distinct cell classes and 86 cell types/states, including diverse populations of astrocytes, oligodendrocytes, oligodendrocyte precursor cells, microglia, and macrophages. Furthermore, snATAC-seq profiling of approximately 1 million nuclei generated high-resolution chromatin accessibility landscapes, enabling the identification of cis-regulatory elements and transcription factors specific to individual cell classes and types. As part of the Human Cell Atlas initiative, our ON atlas serves as a crucial resource for studying optic nerve physiology and deepening our understanding of the cellular and molecular mechanisms underlying optic neuropathies and related diseases.
Datasets (5)
snRNA-seq of human optic nerve and optic nerve head - endothelial cells34,934 cells
snRNA-seq of human optic nerve and optic nerve head - fibroblasts118,925 cells
snRNA-seq of human optic nerve and optic nerve head - oligodendrocytes278,040 cells
snRNA-seq of human optic nerve and optic nerve head - astrocytes273,711 cells
snRNA-seq of human optic nerve and optic nerve head - all cells959,629 cells
The liver is the largest solid organ in the body, yet it remains incompletely characterized. Here we present a spatial proteogenomic atlas of the healthy and obese human and murine liver combining single-cell CITE-seq, single-nuclei sequencing, spatial transcriptomics, and spatial proteomics. By integrating these multi-omic datasets, we provide validated strategies to reliably discriminate and localize all hepatic cells, including a population of lipid-associated macrophages (LAMs) at the bile ducts. We then align this atlas across seven species, revealing the conserved program of bona fide Kupffer cells and LAMs. We also uncover the respective spatially resolved cellular niches of these macrophages and the microenvironmental circuits driving their unique transcriptomic identities. We demonstrate that LAMs are induced by local lipid exposure, leading to their induction in steatotic regions of the murine and human liver, while Kupffer cell development crucially depends on their cross-talk with hepatic stellate cells via the evolutionarily conserved ALK1-BMP9/10 axis.
Datasets (4)
CD45 negative cells from human liver dataset15,481 cells
Myeloid cells from human liver dataset40,821 cells
Lymphoid cells from human liver dataset105,790 cells
Fibroblasts are essential cells that sculpt the architecture and cellular microenvironments within various tissues. Understanding fibroblast heterogeneity and their spatial context in health and disease has major clinical relevance. In this study, we constructed a spatially-resolved atlas of 357 276 human skin fibroblasts from healthy skin and 23 skin disorders. We define 6 major skin fibroblast populations in health and a further three skin disease-specific fibroblast subtypes, and demonstrate the fibroblast composition in different types of skin disease. We also show that inflammatory myofibroblasts (IL11+MMP1+CXCL5+IL7R+) present in early human wounds are a conserved fibroblast subtype in inflammatory disorders and cancer, that likely recruit neutrophils and monocytes to different human tissues. We characterise a human-specific fibroblastic reticular cell (FRC)-like subtype in the skin superficial perivascular niche and postulate their origin from prenatal skin lymphoid tissue organiser (LTo)-like cells. We provide a harmonised nomenclature for skin fibroblasts that integrates previous findings from human skin and other tissues.
Datasets (2)
Skin fibroblast scRNA-seq atlas in health153,546 cells
Skin fibroblasts - Pan-disease fibroblast atlas in skin fibrosis, inflammation, and cancer337,376 cells
Organoids of the endoderm can recapitulate aspects of developing and adult human physiology. Organoids derived from embryonic or induced pluripotent stem cells model development and are guided to specific tissue types via morphogens, whereas organoids derived from tissue-resident fetal or adult stem cells are organ-identity-determined and may model ontogenetic features. However, it has remained difficult to assess the similarity and differences between organoid protocols, and to understand the precision and accuracy of organoid cell states through comparison with primary counterparts. Advances in computational single-cell biology allow the comprehensive integration of datasets with high technical variability. Here, we integrate published single-cell transcriptome datasets from organoids of diverse endoderm-derived tissues including lung, pancreas, intestine, salivary glands, liver, biliary system, stomach, and prostate to establish an initial version of a human endoderm organoid cell atlas (HEOCA). The integration includes nearly one million cells across diverse conditions and data sources. We align and compare cell types and states between organoid models, and harmonize cell type annotations by mapping the atlas to primary tissue counterparts. We focus on intestine and lung, and clarify developmental and adult physiology that can be modeled in vitro. We provide examples of data incorporation from new organoid protocols to expand the atlas, and showcase how comparison to the atlas can illuminate interesting biological features of new datasets. We also show that mapping disease organoid single-cell samples to HEOCA identifies shifts in cell proportion and gene expressions between normal and diseased cells. Taken together, the atlas makes diverse datasets centrally available, and it will be useful to assess organoid fidelity, characterize perturbed and diseased states, streamline protocol development, and will continuously grow in the future.
Understanding kidney health and disease relies upon defining the complexity of cell types and states, their associated molecular profiles, and interactions within tissue neighborhoods. We have applied single-cell and single-nucleus assays to a broad spectrum of healthy reference and disease kidneys. This has provided a high-resolution cellular atlas that includes rare and novel cell populations. We further identify and define cellular states altered in kidney injury, encompassing cycling, adaptive or maladaptive repair, transitioning and degenerative states affecting several nephron segments. These analyses further define biological pathways relevant to injury niches, including signatures underlying the transition from reference to predicted maladaptive states that are associated with a decline in kidney function during chronic kidney disease. Our collaborative efforts across the Kidney Precision Medicine Project (KPMP), the Human BioMolecular Atlas Program (HuBMAP) and the Human Cell Atlas consortia aim to generate a comprehensive human kidney cell atlas that includes injury and disease-relevant cell states and neighborhoods as a valuable resource to the research community. Here we provide an expanded data set from that published previously (Lake et al., Nature 2023) that includes additional KPMP biopsy data.
Datasets (2)
Single-nucleus RNA-seq of the Adult Human Kidney (Version 1.5)304,989 cells
Single-cell RNA-seq of the Adult Human Kidney (Version 1.5)225,177 cells
The Human Cell Atlas data provides the “recipe” for tissue engineering, including organoids. Human organoids – three-dimensional structures of cells that recapitulate organ development in vitro – hold tremendous potential for biomedical applications. They provide tractable models of human physiology and pathology, thereby enabling interventional studies that are difficult or impossible to conduct in human subjects, and reducing the need for animal experiments. Moreover, they can provide patient-specific “avatars” for drug development and personalized therapies, and they help advance regenerative medicine, with the ultimate goal of aiming towards regenerative medicine, and producing functional biological structures that can be transplanted into patients.
Single-cell epigenome/transcriptome sequencing and spatial profiling can provide a comprehensive assessment of cell composition and cell states within organoids, enabling a comparison to matched primary tissue samples. These comparisons can catalyse the development of better organoid protocols. The Organoid Cell Atlas will foster the production, quality control, dissemination, and utilization of single-cell data for human organoids, and it will connect such datasets with the comprehensive profiles of primary tissue that are being generated within the HCA.
The trabecular meshwork (TM) and ciliary body (CB) play critical roles in regulating aqueous humor homeostasis in the human eye. The TM serves as the primary outflow pathway for aqueous humor, while the CB controls lens shape and produces aqueous humor. Both tissues are composed of diverse cell types, with cell abundances varying across several orders of magnitude. To create a comprehensive single-cell atlas and advance our understanding of TM and CB biology, we generated large-scale datasets from healthy human donors using single-nucleus RNA-seq (snRNA-seq), single-cell RNA-seq (scRNA-seq), and single-nucleus ATAC-seq (snATAC-seq) across a range of ages, genders, and ethnic backgrounds. Additionally, we performed a meta-analysis by integrating previously published datasets. All data were uniformly preprocessed and combined, resulting in a comprehensive TM and CB atlas that includes transcriptomic profiles of over 1 million cells from 70 donors. We identified 19 distinct cell types, including rare populations such as immune cells, with the least abundant types representing as little as 0.08% of the total cell population. Furthermore, snATAC-seq profiling of over 500,000 nuclei provided detailed chromatin accessibility landscapes, allowing for the identification of cis-regulatory elements specific to each cell type. All datasets are publicly accessible via CELLxGENE and the UCSC Cell Browser, enabling interactive exploration of gene expression and chromatin states at the single-cell level. As part of the Human Cell Atlas initiative, our TM and CB atlas offers a valuable resource for understanding eye physiology and establishes a foundation for further research into the cellular and molecular mechanisms regulating intraocular pressure and related ocular diseases.
Datasets (2)
Single cell atlas of the human trabecular meshwork and ciliary body - scRNA-seq332,995 cells
Single cell atlas of the human trabecular meshwork and ciliary body - snRNA-seq1,102,250 cells
As the light sensing part of the visual system, the human retina is composed of five classes of neuron, including photoreceptors, horizontal cells, amacrine, bipolar, and retinal ganglion cells. Each class of neuron can be further classified into subgroups with the abundance varying three orders of magnitude. Therefore, to capture all cell types in the retina and generate a complete single cell reference atlas, it is essential to scale up from currently published single cell profiling studies to improve the sensitivity. In addition, to gain a better understanding of gene regulation at single cell level, it is important to include sufficient scATAC-seq data in the reference. To fill the gap, we performed snRNA-seq and snATAC-seq for the retina from healthy donors. To further increase the size of the dataset, we then collected and incorporated publicly available datasets. All data underwent a unified preprocessing pipeline and data integration. Multiple integration methods were benchmarked by scIB, and scVI was chosen. To harness the power of multiomics, snATAC-seq datasets were also preprocessed, and scGlue was used to generate co-embeddings between snRNA-seq and snATAC-seq cells. To facilitate the public use of references, we employ CELLxGENE and UCSC Cell Browser for visualization. By combining previously published and newly generated datasets, a single cell atlas of the human retina that is composed of 2.5 million single cells from 48 donors has been generated. As a result, over 90 distinct cell types are identified based on the transcriptomics profile with the rarest cell type accounting for about 0.01% of the cell population. In addition, open chromatin profiling has been generated for over 400K nuclei via single nuclei ATAC-seq, allowing systematic characterization of cis-regulatory elements for individual cell type. Integrative analysis reveals intriguing differences in the transcriptome, chromatin landscape, and gene regulatory network among cell class, subgroup, and type. In addition, changes in cell proportion, gene expression and chromatin openness have been observed between different gender and over age. Accessible through interactive browsers, this study represents the most comprehensive reference cell atlas of the human retina to date. As part of the human cell atlas project, this resource lays the foundation for further research in understanding retina biology and diseases.
Datasets (6)
Single cell atlas of the human retina - Bipolar cell subclass - scRNA-seq72,788 cells
Single cell atlas of the human retina - all cells - scRNA-seq265,767 cells
snRNA-seq of human retina - retinal ganglion cell subset399,605 cells
snRNA-seq of human retina - amacrine cell subset571,579 cells
snRNA-seq of human retina - bipolar cell subset691,008 cells
snRNA-seq of human retina - all cells3,177,310 cells
Tonsils are constantly exposed to antigens through the upper respiratory tract, making them a valuable secondary lymphoid organ (SLO) for studying the interaction between innate and adaptive immune cells during germinal center (GC) development, which is crucial for building adaptive immunity. This reference includes 377,963 cells (10x Genomics 3' v3) from 17 healthy human donors spanning various age groups, including children, young adults, and older adults.
The annotation provided in this first version comprises 42 categories that provide a stable categories to classify single-cell transcriptomes of SLOs, useful for tools like Azimuth (see external URLs). In the next version, we will add a more detailed classification, encompassing all cell types and states identified in the tonsil atlas. A validation cohort was included to confirm the presence and accuracy of each annotation this latter level, using criteria such as cell neighborhood preservation, conservation of bona fide marker genes, and annotation confidence derived from the KNN classifier. For a full description and interpretation of this validation cohort, please refer to the final section of the manuscript.
The tonsil atlas is a FAIR resource—findable, accessible, interoperable, and reusable. The raw data has been deposited in ArrayExpress and can be remapped and reanalyzed by following the instructions provided in the TonsilAtlas and TonsilAtlasCAP GitHub repositories. The resulting expression matrices and Seurat objects are available on Zenodo, and the data can be accessed, explored, interpreted, and reused via the HCATonsilAtlas Bioconductor package and the Azimuth web interface. Additionally, we provide a detailed glossary of the marker genes, rationales, and publications used to annotate each cell type and state. The external URLs section contains links to all these resources.
Datasets (1)
An Atlas of Cells in the Human Tonsil377,963 cells
Crohn’s disease is an inflammatory bowel disease (IBD) commonly treated through anti-TNF blockade. However, most patients still relapse and inevitably progress. Comprehensive single-cell RNA-sequencing (scRNA-seq) atlases have largely sampled patients with established treatment-refractory IBD, limiting our understanding of which cell types, subsets, and states at diagnosis anticipate disease severity and response to treatment. Here, through combining clinical, flow cytometry, histology, and scRNA-seq methods, we profile diagnostic human biopsies from the terminal ileum of treatment-naïve pediatric patients with Crohn’s disease (pediCD; n=14), matched repeat biopsies (pediCD-treated; n=8) and from non-inflamed pediatric controls with functional gastrointestinal disorders (FGID; n=13). To resolve and annotate epithelial, stromal, and immune cell states among the 201,883 baseline single-cell transcriptomes, we develop a principled and unbiased tiered clustering approach, ARBOL. Through flow cytometry and scRNA-seq, we observe that treatment-naïve pediCD and FGID have similar broad cell type composition. However, through high-resolution scRNA-seq analysis and microscopy, we identify significant differences in cell subsets and states that arise during pediCD relative to FGID. By closely linking our scRNA-seq analysis with clinical meta-data, we resolve a vector of T cell, innate lymphocyte, myeloid, and epithelial cell states in treatment-naïve pediCD (pediCD-TIME) samples which can distinguish patients along the trajectory of disease severity and anti-TNF response. By using ARBOL with integration, we position repeat on-treatment biopsies from our patients between treatment-naïve pediCD and on-treatment adult CD. We identify that anti-TNF treatment pushes the pediatric cellular ecosystem towards an adult, more treatment-refractory state. Our study jointly leverages a treatment-naïve cohort, high-resolution principled scRNA-seq data analysis, and clinical outcomes to understand which baseline cell states may predict Crohn’s disease trajectory.
Datasets (1)
Concerted changes in the pediatric single-cell intestinal ecosystem before and after anti-TNF blockade197,281 cells
Lack of diversity and proportionate representation in genomics datasets and databases contributes to inequity in healthcare outcomes globally. The relationships of human diversity with biological and biomedical phenotypes are pervasive, yet remain understudied, particularly in a single-cell genomics context. Here we present the Asian Immune Diversity Atlas (AIDA), a multi-national single-cell RNA-sequencing (scRNA-seq) healthy reference atlas of human immune cells. AIDA comprises 1,265,624 circulating immune cells from 619 healthy donors and 6 controls, spanning 7 population groups across 5 countries. AIDA is one of the largest healthy blood datasets in terms of number of cells, and also the most diverse in terms of number of population groups. Though population groups are frequently compared at the continental level, we identified a pervasive impact of sub-continental diversity on cellular and molecular properties of immune cells. These included cell populations and genes implicated in disease risk and pathogenesis as well as those relevant for diagnostics. We identified numerous examples where the effects of age and sex were modulated by self-reported ethnicity. We also detected age, sex, and self-reported ethnicity differences at the resolution of cell neighbourhoods, highlighting finer-grained distinctions than were apparent at cell-type level. We discovered functional genetic variants influencing cell type-specific gene expression, including context-dependent effects, which were under-represented in analyses of non-Asian population groups, and which helped contextualise disease-associated variants. We validated our findings using multiple independent datasets and cohorts. AIDA provides fundamental insights into the relationships of human diversity with immune cell phenotypes, enables analyses of multi-ancestry disease datasets, and facilitates the development of precision medicine efforts in Asia and beyond.
Datasets (1)
AIDA Phase 1 Data Freeze v2: Chinese, Indian, Japanese, Korean, Malay, and Thai donors in Japan, Singapore, South Korea, Thailand, and India1,265,624 cells
