Abstract
Cellular diversity is governed not only by the transcriptome but also by multiple layers of epigenomic regulation, including nucleosome occupancy, chromatin states and genome architecture1,2,3. Here, to comprehensively understand how these regulatory modalities converge to shape cellular identity, we developed a single-cell four-omics sequencing method that enables parallel profiling of genome conformation, histone modifications, chromatin accessibility and gene expression within the same cell (CHARM). Applying CHARM to mouse embryonic stem cells and cortical tissues, we reconstructed integrated epigenome profiles, uncovering distinct cell-cycle dynamics of chromatin accessibility and histone modification, and spatial clustering of regulatory elements in three-dimensional nuclear space. Leveraging an interpretable machine learning model, we further identified thousands of enhancer–promoter linkages with high accuracy that modulate gene expression in a cell-type- and subtype-specific manner. Together, CHARM enables integrative dissection of the three-dimensional epigenome at single-cell resolution, providing a versatile platform for decoding the regulatory landscape across diverse cells in complex tissues.
This is a preview of subscription content, access via your institution
Access options
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$32.99 / 30 days
cancel any time
Subscribe to this journal
Receive 51 print issues and online access
$199.00 per year
only $3.90 per issue
Buy this article
- Purchase on SpringerLink
- Instant access to the full article PDF.
USD 39.95
Prices may be subject to local taxes which are calculated during checkout
Similar content being viewed by others
Data availability
Raw sequencing data generated in this study have been deposited in the NCBI Sequence Read Archive (SRA) under BioProject accession PRJNA1284811. Processed and analysed datasets have been deposited in the NCBI Gene Expression Omnibus (GEO) under accession GSE303006. The mouse reference genome assembly and gene annotation used in this study were GRCm38 with GENCODE release M23 (https://www.gencodegenes.org/mouse/release_M23.html). Phased mouse SNPs were obtained from the Sanger Institute Mouse Genomes Project (https://ftp.ebi.ac.uk/pub/databases/mousegenomes/REL-1505-SNPs_Indels/mgp.v5.merged.indels.dbSNP142.normed.vcf.gz). The mouse brain single-cell ATAC–seq data were obtained from the CELLxGENE data portal (https://cellxgene.cziscience.com/collections/5e469121-c203-4775-962d-dcf2e5d6a472). The mouse brain cortex single-cell RNA atlas were obtained from Allen Institute (https://idk-etl-prod-download-bucket.s3.amazonaws.com/aibs_mouse_ctx-hpf_smart-seq/Seurat.ss.rda). The mESC droplet-based paired-tag processed H3K27ac data were obtained from GEO (GSE224560). Human genetic variants associated with intelligence were obtained from the NHGRI-EBI GWAS Catalog (trait accession EFO_0004337).
Code availability
Code for data preprocessing and analysis is available at https://github.com/skelviper/CHARM. Code and 3D-printable models for the automated liquid-handling workflow used during library preparation are available at https://github.com/skelviper/OT2.
References
Klemm, S. L., Shipony, Z. & Greenleaf, W. J. Chromatin accessibility and the regulatory epigenome. Nat. Rev. Genet. 20, 207–220 (2019).
Kouzarides, T. Chromatin modifications and their function. Cell 128, 693–705 (2007).
Misteli, T. The self-organizing genome: principles of genome architecture and function. Cell 183, 28–45 (2020).
Buenrostro, J. D., Giresi, P. G., Zaba, L. C., Chang, H. Y. & Greenleaf, W. J. Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin, DNA-binding proteins and nucleosome position. Nat. Methods 10, 1213–1218 (2013).
Kaya-Okur, H. S. et al. CUT&Tag for efficient epigenomic profiling of small samples and single cells. Nat. Commun. 10, 1930 (2019).
Lieberman-Aiden, E. et al. Comprehensive mapping of long-range interactions reveals folding principles of the human genome. Science 326, 289–293 (2009).
Baysoy, A., Bai, Z., Satija, R. & Fan, R. The technological landscape and applications of single-cell multi-omics. Nat. Rev. Mol. Cell Biol. 24, 695–713 (2023).
Chappell, L., Russell, A. J. C. & Voet, T. Single-cell (multi)omics technologies. Annu. Rev. Genomics Hum. Genet. 19, 15–41 (2018).
Zhu, C., Preissl, S. & Ren, B. Single-cell multimodal omics: the power of many. Nat. Methods 17, 11–14 (2020).
Lee, D. S. et al. Simultaneous profiling of 3D genome structure and DNA methylation in single human cells. Nat. Methods 16, 999–1006 (2019).
Li, G. et al. Joint profiling of DNA methylation and chromatin architecture in single cells. Nat. Methods 16, 991–993 (2019).
Liu, Z. et al. Linking genome structures to functions by simultaneous single-cell Hi-C and RNA-seq. Science 380, 1070–1076 (2023).
Qu, J. et al. Simultaneous profiling of chromatin architecture and transcription in single cells. Nat. Struct. Mol. Biol. 30, 1393–1402 (2023).
Wen, X. et al. Single-cell multiplex chromatin and RNA interactions in ageing human brain. Nature 628, 648–656 (2024).
Wu, H. et al. Simultaneous single-cell three-dimensional genome and gene expression profiling uncovers dynamic enhancer connectivity underlying olfactory receptor choice. Nat. Methods 21, 974–982 (2024).
Zhou, T. et al. GAGE-seq concurrently profiles multiscale 3D genome organization and gene expression in single cells. Nat. Genet. 56, 1701–1711 (2024).
Chang, L. et al. Droplet Hi-C enables scalable, single-cell profiling of chromatin architecture in heterogeneous tissues. Nat. Biotechnol. 43, 1694–1707 (2025).
Chai, H. et al. Tri-omic single-cell mapping of the 3D epigenome and transcriptome in whole mouse brains throughout the lifespan. Nat. Methods 22, 994–1007 (2025).
Wu, H., Wang, M., Zheng, Y. & Xie, X. S. Droplet-based high-throughput 3D genome structure mapping of single cells with simultaneous transcriptomics. Cell Discov. 11, 8 (2025).
Luecken, M. D. et al. Benchmarking atlas-level data integration in single-cell genomics. Nat. Methods 19, 41–50 (2022).
Hu, Y. et al. Benchmarking algorithms for single-cell multi-omics prediction and integration. Nat. Methods 21, 2182–2194 (2024).
Fu, S. et al. Benchmarking single-cell multi-modal data integrations. Nat. Methods 22, 2437–2448 (2025).
Flynn, E., Almonte-Loya, A. & Fragiadakis, G. K. Single-cell multiomics. Annu. Rev. Biomed. Data Sci. 6, 313–337 (2023).
Cao, J. et al. Joint profiling of chromatin accessibility and gene expression in thousands of single cells. Science 361, 1380–1385 (2018).
Chen, S., Lake, B. B. & Zhang, K. High-throughput sequencing of the transcriptome and chromatin accessibility in the same cell. Nat. Biotechnol. 37, 1452–1457 (2019).
Zhu, C. et al. An ultra high-throughput method for single-cell joint analysis of open chromatin and transcriptome. Nat. Struct. Mol. Biol. 26, 1063–1070 (2019).
Ma, S. et al. Chromatin potential identified by shared single-cell profiling of RNA and chromatin. Cell 183, 1103–1116.e20 (2020).
Tan, L. et al. Changes in genome architecture and transcriptional dynamics progress independently of sensory experience during post-natal brain development. Cell 184, 741–758.e17 (2021).
Xiong, H., Luo, Y., Wang, Q., Yu, X. & He, A. Single-cell joint detection of chromatin occupancy and transcriptome enables higher-dimensional epigenomic reconstructions. Nat. Methods 18, 652–660 (2021).
Xu, W. et al. ISSAAC-seq enables sensitive and flexible multimodal profiling of chromatin accessibility and gene expression in single cells. Nat. Methods 19, 1243–1249 (2022).
Bartosovic, M. & Castelo-Branco, G. Multimodal chromatin profiling using nanobody-based single-cell CUT&Tag. Nat. Biotechnol. 41, 794–805 (2023).
Xie, Y. et al. Droplet-based single-cell joint profiling of histone modifications and transcriptomes. Nat. Struct. Mol. Biol. 30, 1428–1433 (2023).
Sarkar, A. & Hochedlinger, K. The Sox family of transcription factors: versatile regulators of stem and progenitor cell fate. Cell Stem Cell 12, 15–30 (2013).
Li, Y. et al. CRISPR reveals a distal super-enhancer required for Sox2 expression in mouse embryonic stem cells. PLoS ONE 9, e114485 (2014).
Alexamder, J. M. et al. Live-cell imaging reveals enhancer-dependent Sox2 transcription in the absence of enhancer proximity. eLife 8, e41769 (2019).
Di Croce, L. & Helin, K. Transcriptional regulation by Polycomb group proteins. Nat. Struct. Mol. Biol. 20, 1147–1155 (2013).
Wang, M. & Zhang, Y. Tn5 transposase-based epigenomic profiling methods are prone to open chromatin bias. Preprint at bioRxiv https://doi.org/10.1101/2021.07.09.451758 (2021).
Bernstein, B. E. et al. A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell 125, 315–326 (2006).
Macrae, T. A., Fothergill-Robinson, J. & Ramalho-Santos, M. Regulation, functions and transmission of bivalent chromatin during mammalian development. Nat. Rev. Mol. Cell Biol. 24, 6–26 (2023).
Stewart-Morgan, K. R., Petryk, N. & Groth, A. Chromatin replication and epigenetic cell memory. Nat. Cell Biol. 22, 361–371 (2020).
Nagano, T. et al. Cell-cycle dynamics of chromosomal organization at single-cell resolution. Nature 547, 61–67 (2017).
Mahat, D. B. et al. Single-cell nascent RNA sequencing unveils coordinated global transcription. Nature 631, 216–223 (2024).
Pope, B. D. et al. Topologically associating domains are stable units of replication-timing regulation. Nature 515, 402–405 (2014).
Owen, J. A., Osmanovic, D. & Mirny, L. Design principles of 3D epigenetic memory systems. Science 382, eadg3053 (2023).
Nagano, T. et al. Single-cell Hi-C reveals cell-to-cell variability in chromosome structure. Nature 502, 59–64 (2013).
Stevens, T. J. et al. 3D structures of individual mammalian genomes studied by single-cell Hi-C. Nature 544, 59–64 (2017).
Tan, L., Xing, D., Chang, C. H., Li, H. & Xie, X. S. Three-dimensional genome structures of single diploid human cells. Science 361, 924–928 (2018).
Lafontaine, D. L., Yang, L., Dekker, J. & Gibcus, J. H. Hi-C 3.0: improved protocol for genome-wide chromosome conformation capture. Curr. Protoc. 1, e198 (2021).
Wu, H., Zhang, J., Tan, L. & Xie, X. S. Single-cell Micro-C profiles 3D genome structures at high resolution and characterizes multi-enhancer hubs. Nat. Genet. 57, 1777–1786 (2025).
Caetano, F. A. et al. MIiSR: molecular interactions in super-resolution imaging enables the analysis of protein interactions, dynamics and formation of multi-protein structures. PLoS Comput. Biol. 11, e1004634 (2015).
Vieux-Rochas, M., Fabre, P. J., Leleu, M., Duboule, D. & Noordermeer, D. Clustering of mammalian Hox genes with other H3K27me3 targets within an active nuclear domain. Proc. Natl Acad. Sci. USA 112, 4672–4677 (2015).
Xie, L. et al. 3D ATAC-PALM: super-resolution imaging of the accessible genome. Nat. Methods 17, 430–436 (2020).
Bintu, B. et al. Super-resolution chromatin tracing reveals domains and cooperative interactions in single cells. Science 362, eaau1783 (2018).
Whyte, W. A. et al. Master transcription factors and mediator establish super-enhancers at key cell identity genes. Cell 153, 307–319 (2013).
Cho, W. K. et al. Mediator and RNA polymerase II clusters associate in transcription-dependent condensates. Science 361, 412–415 (2018).
Sabari, B. R. Coactivator condensation at super-enhancers links phase separation and gene control. Science 361, eaar3958 (2018).
Lim, B. & Levine, M. S. Enhancer-promoter communication: hubs or loops?. Curr. Opin. Genet. Dev. 67, 5–9 (2021).
Uyehara, C. M. & Apostolou, E. 3D enhancer-promoter interactions and multi-connected hubs: organizational principles and functional roles. Cell Rep. 42, 112068 (2023).
Hao, Y. et al. Integrated analysis of multimodal single-cell data. Cell 184, 3573–3587.e29 (2021).
Mitra, S. et al. Single-cell multi-ome regression models identify functional and disease-associated enhancers and enable chromatin potential analysis. Nat. Genet. 56, 627–636 (2024).
Fulco, C. P. et al. Activity-by-contact model of enhancer-promoter regulation from thousands of CRISPR perturbations. Nat. Genet. 51, 1664–1669 (2019).
Savage, J. E. et al. Genome-wide association meta-analysis in 269,867 individuals identifies new genetic and functional links to intelligence. Nat. Genet. 50, 912–919 (2018).
Whitton, L. et al. Genes regulated by SATB2 during neurodevelopment contribute to schizophrenia and educational attainment. PLoS Genet. 14, e1007515 (2018).
Cera, I. et al. Genes encoding SATB2-interacting proteins in adult cerebral cortex contribute to human cognitive ability. PLoS Genet. 15, e1007890 (2019).
Zarate, Y. A. et al. Natural history and genotype-phenotype correlations in 72 individuals with SATB2-associated syndrome. Am. J. Med. Genet. A 176, 925–935 (2018).
Jaitner, C. et al. Satb2 determines miRNA expression and long-term memory in the adult central nervous system. eLife 5, e17361 (2016).
Li, Y. et al. Satb2 ablation impairs hippocampus-based long-term spatial memory and short-term working memory and immediate early genes (IEGs)-mediated hippocampal synaptic plasticity. Mol. Neurobiol. 43, 15251–15266 (2025).
Wahl, N. et al. SATB2 organizes the 3D genome architecture of cognition in cortical neurons. Mol. Cell 84, 621–639.e9 (2024).
Schep, A. N., Wu, B., Buenrostro, J. D. & Greenleaf, W. J. chromVAR: inferring transcription-factor-associated accessibility from single-cell epigenomic data. Nat. Methods 14, 975–978 (2017).
Pampari, A. et al. ChromBPNet: bias factorized, base-resolution deep learning models of chromatin accessibility reveal cis-regulatory sequence syntax, transcription factor footprints and regulatory variants. Preprint at bioRxiv https://doi.org/10.1101/2024.12.25.630221 (2024).
Reveron-Gomez, N. et al. Accurate recycling of parental histones reproduces the histone modification landscape during DNA replication. Mol. Cell 72, 239–249.e5 (2018).
Fennessy, R. T. & Owen-Hughes, T. Establishment of a promoter-based chromatin architecture on recently replicated DNA can accommodate variable inter-nucleosome spacing. Nucleic Acids Res. 44, 7189–7203 (2016).
Ramachandran, S. & Henikoff, S. Transcriptional regulators compete with nucleosomes post-replication. Cell 165, 580–592 (2016).
Stewart-Morgan, K. R., Reveron-Gomez, N. & Groth, A. Transcription restart establishes chromatin accessibility after DNA replication. Mol. Cell 75, 408–414 (2019).
Blayney, J. W. et al. Super-enhancers include classical enhancers and facilitators to fully activate gene expression. Cell 186, 5826–5839.e18 (2023).
Mazzocca, M., Narducci, D. N., Grosse-Holz, S., Matthias, J. & Hansen, A. S. Chromatin dynamics are highly subdiffusive across seven orders of magnitude. Preprint at bioRxiv https://doi.org/10.1101/2025.05.10.653248 (2025).
Sanyal, A., Lajoie, B. R., Jain, G. & Dekker, J. The long-range interaction landscape of gene promoters. Nature 489, 109–113 (2012).
Tian, W. et al. Single-cell DNA methylation and 3D genome architecture in the human brain. Science 382, eadf5357 (2023).
Liu, H. et al. Single-cell DNA methylome and 3D multi-omic atlas of the adult mouse brain. Nature 624, 366–377 (2023).
Heffel, M. G. et al. Temporally distinct 3D multi-omic dynamics in the developing human brain. Nature 635, 481–489 (2024).
Macosko, E. Z. et al. Highly parallel genome-wide expression profiling of individual cells using nanoliter droplets. Cell 161, 1202–1214 (2015).
Klein, A. M. et al. Droplet barcoding for single-cell transcriptomics applied to embryonic stem cells. Cell 161, 1187–1201 (2015).
Vitak, S. A. et al. Sequencing thousands of single-cell genomes with combinatorial indexing. Nat. Methods 14, 302–308 (2017).
Rosenberg, A. B. et al. Single-cell profiling of the developing mouse brain and spinal cord with split-pool barcoding. Science 360, 176–182 (2018).
Meers, M. P., Llagas, G., Janssens, D. H., Codomo, C. A. & Henikoff, S. Multifactorial profiling of epigenetic landscapes at single-cell resolution using MulTI-Tag. Nat. Biotechnol. 41, 708–716 (2023).
Gopalan, S., Wang, Y., Harper, N. W., Garber, M. & Fazzio, T. G. Simultaneous profiling of multiple chromatin proteins in the same cells. Mol. Cell 81, 4736–4746.e5 (2021).
Stoeckius, M. et al. Simultaneous epitope and transcriptome measurement in single cells. Nat. Methods 14, 865–868 (2017).
Mimitou, E. P. et al. Scalable, multimodal profiling of chromatin accessibility, gene expression and protein levels in single cells. Nat. Biotechnol. 39, 1246–1258 (2021).
Meissner, A., Eminli, S. & Jaenisch, R. Derivation and manipulation of murine embryonic stem cells. Methods Mol. Biol. 482, 3–19 (2009).
Hao, Y. et al. Dictionary learning for integrative, multimodal and scalable single-cell analysis. Nat. Biotechnol. 42, 293–304 (2024).
Stuart, T., Srivastava, A., Madad, S., Lareau, C. A. & Satija, R. Single-cell chromatin state analysis with Signac. Nat. Methods 18, 1333–1341 (2021).
Zhou, J. et al. Robust single-cell Hi-C clustering by convolution- and random-walk-based imputation. Proc. Natl Acad. Sci. USA 116, 14011–14018 (2019).
Bonev, B. et al. Multiscale 3D genome rewiring during mouse neural development. Cell 171, 557–572.e24 (2017).
Cai, Y. et al. H3K27me3-rich genomic regions can function as silencers to repress gene expression via chromatin interactions. Nat. Commun. 12, 719 (2021).
Hinrichs, A. S. et al. The UCSC Genome Browser Database: update 2006. Nucleic Acids Res. 34, D590–D598 (2006).
Murphy, D. et al. 3D Enhancer-promoter networks provide predictive features for gene expression and coregulation in early embryonic lineages. Nat. Struct. Mol. Biol. 31, 125–140 (2024).
Agrawal, P. et al. Genome editing demonstrates that the -5 kb Nanog enhancer regulates Nanog expression by modulating RNAPII initiation and/or recruitment. J. Biol. Chem. 296, 100189 (2021).
Tesar, P. J. et al. New cell lines from mouse epiblast share defining features with human embryonic stem cells. Nature 448, 196–199 (2007).
Xie, L. et al. A dynamic interplay of enhancer elements regulates Klf4 expression in naive pluripotency. Genes Dev. 31, 1795–1808 (2017).
Engreitz, J. M. et al. Local regulation of gene expression by lncRNA promoters, transcription and splicing. Nature 539, 452–455 (2016).
Moorthy, S. D. et al. Enhancers and super-enhancers have an equivalent regulatory role in embryonic stem cells through regulation of single or multiple genes. Genome Res. 27, 246–258 (2017).
Acknowledgements
This work is supported by National Natural Science Foundation of China (32430061, 324B200071 and T2225001), Noncommunicable Chronic Diseases-National Science and Technology Major Project (2023ZD0501200), National Key R&D Program of China (2021ZD0202502) and Beijing Advanced Innovation Center for Genomics. We thank Z. Liu, J. Yu and M. Wang for helpful discussion, and the Protein Preparation and Identification Core at National Center for Protein Sciences at Peking University for help with cell sorting and sequencing.
Author information
Authors and Affiliations
Contributions
Y. Chen, Z.L. and D.X. designed the experiments. Y. Chen performed the CHARM experiments and data collection. Z.L. and D.X. analysed the data. H.X. performed the animal tissue collection. B.L. helped with experiments. J.L., M.W., Y. Chi, M.L., Y.P. and H.G. helped with data analysis. Y. Chen, Z.L. and D.X. prepared the manuscript; D.X. managed and supervised the project.
Corresponding author
Ethics declarations
Competing interests
D.X., Y. Chen and Z.L. are inventors on a patent that covers CHARM. The other authors declare no competing interests.
Peer review
Peer review information
Nature thanks Sheng Zhong and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data figures and tables
Extended Data Fig. 1 Quality control and benchmarking of CHARM.
a,b, Scatter plots showing quality control metrics used for filtering CHARM data. a, Fraction of reads in peaks (FRiP) for H3K27me3 versus chromatin accessibility. b, Number of Hi-C contacts versus RNA UMIs per cell. Dashed lines indicate filtering thresholds. c, Violin plots showing TSS enrichment per cell across different methods. d-g, Violin plots comparing per-cell data yield from CHARM with other single-cell methods across four modalities: d, RNA UMIs; e, H3K27me3 fragments; f, accessible chromatin fragments; and g, Hi-C contacts. Boxes indicate the median and 25th-75th percentiles, with whiskers extending to 1.5× the interquartile range in panel c-g. h, Scatter plot of Sox2 UMI counts versus enhancer–promoter (E–P) interaction strength, with cells ordered by Sox2 expression. Each dot represents one cell (n = 720 cells). Pearson’s correlation coefficient is indicated. i, Box plots comparing E–P interaction strength between cells with high and low Sox2 expression. Cells were divided into high (n = 176 cells) and low (n = 544 cells) Sox2 expression groups. Boxes indicate the median and 25th-75th percentiles, with whiskers extending to 1.5× the interquartile range. Statistical significance was assessed using a two-sided Wilcoxon rank-sum test (P = 0.72).
Extended Data Fig. 2 CHARM shows minimal open-chromatin bias and enables single-cell analysis of bivalent chromatin.
a, Pileup plot (top) and heatmap (bottom) showing chromatin accessibility signals at the TSS of expressed and unexpressed genes. b, Pileup plot (top) and heatmap (bottom) showing H3K27me3 signals at the TSS of expressed and unexpressed genes. c, Pileup plot (top) and heatmap (bottom) showing H3K27me3 signals at bivalent and silenced (neither expressed nor bivalent) promoters. d, Histogram of the odds ratio for co-occurrence of accessibility and H3K27me3 fragments within the same single cell at bivalent regions. e, Example of bivalent chromatin at single-allele resolution at the Gad2 locus. Each row represents an individual cell (n = 805), with the B6 and CAST alleles shown separately. A bin is labeled as “bivalent” when both accessibility and H3K27me3 signals are detected on the same allele within the same cell.
Extended Data Fig. 3 Distinct restoration kinetics of chromatin accessibility and H3K27me3 across the cell cycle.
a, Schematics illustrating the analytical framework for profiling epigenomic restoration. Left: To assess the reestablishment of chromatin accessibility, cells are ordered along a pseudo-cell-cycle trajectory from M phase—when chromatin is condensed and inaccessible—to G2 phase, reflecting progressive chromatin reopening. Right: For histone mark restoration, cells are ordered from S to G1 phase. A signal-doubling step in M and G1 phases is applied to account for the dilution of H3K27me3 due to cell division. b,c, Percentage of peaks reaching their maximum signal along the pseudo-cell-cycle for chromatin accessibility (b) and H3K27me3 (c). d, Normalized average peak intensity stratified by signal strength, comparing the 1,000 ENCODE H3K27me3 peaks with the highest signal (High H3K27me3) to the 1,000 peaks with the lowest signal (Low H3K27me3). e, Example peak in an early-replicating domain. f, example peak in a late-replicating domain. In both panels, cells are ordered along the cell-cycle trajectory (M to G2). Top tracks show aggregate accessibility; bottom panels display the corresponding single-cell signals. g, Schematic illustrating the prediction of H3K27me3 recovery using 1D or 3D models. LMM, linear mixed-effects model; MSE, mean squared error. h, Box plots showing per-cell mean squared error (MSE) for the 1D and 3D models (n = 193 cells). Boxes indicate the median and 25th-75th percentiles, with whiskers extending to 1.5× the interquartile range. Statistical significance was assessed using a two-sided Wilcoxon signed-rank test (P = 2.06 × 10−31).
Extended Data Fig. 4 Spatial localization of active and repressive chromatin in 3D nuclear space.
a, Classification of A-compartment particles based on z-score normalized accessibility and H3K27me3 fragment counts per particle. Using 0 as the threshold on each axis, particles with accessibility > 0 and H3K27me3 ≤ 0 were labeled accessibility-only (acc-only), those with accessibility ≤ 0 and H3K27me3 > 0 were labeled H3K27me3-only (me3-only), and those with accessibility > 0 and H3K27me3 > 0 were labeled accessibility-plus-H3K27me3 (acc/me3). Particles in the B compartment and remaining particles are grouped into the “Else” category. b, UpSet plot showing the number of ENCODE-defined bivalent regions that overlapping with the particle classifications shown in a. c, Pair cross-correlation C(r) between ENCODE-defined bivalent regions (biv.) and each particle class in a. Values were computed per cell and are shown as the mean across 720 cells with s.e.m.; distance is measured in particle radii (p.r.). d, Pair cross-correlation C(r) between RNA particles and each particle class in a. For panels c, d, e, and g, data are presented as mean values ± s.e.m. (n = 720 cells)., Pair autocorrelation function G(r) for accessibility particles restricted to the A compartments. f, Radial distribution of actual accessibility particles compared to a control set of particles sampled to match a uniform radial distribution (R.G. sampled). g, Pair auto-correlation function G(r) showing that radial distribution minimally affects clustering of accessibility particles. h, Box plots comparing the fraction of accessible particles located within 2D clusters with different stitch sizes versus 3D clusters across all mESC cells (n = 720). i, Waterfall plot showing 2D clusters identified from ranked, stitched accessibility windows. j, Genome browser snapshots of representative 2D clusters defined using various stitch sizes. k, Box plots of per-cell odds ratios showing the enrichment of 3D-clustered particles at all peaks, 2D clustered peaks, or 2D isolated peaks (two-sided Wilcoxon rank-sum test). For panels h and k, boxes indicate the median and 25th-75th percentiles, with whiskers extending to 1.5× the interquartile range (n = 720 cells). l, Histogram comparing the number of distinct 2D clusters contributing to a single 3D cluster in actual versus shuffled datasets. m, Histogram of odds ratios quantifying the enrichment of bulk chromatin loop anchors in 3D accessibility clusters, similar to Fig. 3h. n, Heatmap of aggregated chromatin interaction frequencies between pairs of accessibility particles co-residing in the same 3D cluster. o, Distribution of odds ratios for co-expression among gene pairs within 2-Mb regions that residing in the same 3D accessibility cluster (Actual) compared to shuffled controls. (P = 8.80 × 10−158, ***P < 0.001, two-sided Wilcoxon signed-rank test). p, Distribution of odds ratios for co-expression among gene pairs within the same 2D accessibility cluster (with 50-kb stitch size) that residing in the same 3D accessibility cluster (Actual) compared to shuffled controls. (P = 4.34 × 10−23, ***P < 0.001, two-sided Wilcoxon signed-rank test).
Extended Data Fig. 5 CHARM enables comparative and integrative multi-omics analysis of cell types in the mouse cortex.
a-c, Violin plots comparing per-cell data yield between mESC and mouse cortex datasets for chromatin accessibility (a), chromatin contacts (b), and RNA UMIs (c). 100 randomly sampled cells were analyzed for each dataset. Violin plots show the distribution of per-cell values; embedded box plots indicate the median and 25th-75th percentiles, with whiskers extending to 1.5× the interquartile range. d-f, Heatmaps showing the proportion of cells from RNA-based cell type that overlapped with corresponding clusters defined by chromatin accessibility (d), H3K27ac (e), and chromatin structure (f). g, Box plots of silhouette scores assessing clustering performance based on individual modalities and integrated Weighted Nearest Neighbor (WNN) analysis (n = 3,620 cells). Boxes indicate the median and 25th-75th percentiles, with whiskers extending to 1.5× the interquartile range. Points beyond the whiskers represent outliers. h, Aggregated Hi-C contact maps comparing intratelencephalic and claustrum excitatory neurons (upper right) and other excitatory neuron subtypes (bottom left) generated by Paired Droplet Hi-C, dscHi-C-multiome, and CHARM. The number of cells included in each group is indicated. i, UMAP visualization of the scHiCluster embedding at 500-kb resolution, with cells colored by cell-type labels transferred from a reference RNA atlas to the RNA modality of each dataset using Seurat Canonical Correlation Analysis (CCA). j, UMAP embedding of mouse cortical cells based on integrated WNN analysis. k, Heatmap of modality weights assigned by WNN for each identified cell cluster.
Extended Data Fig. 6 Enrichment of cell type-specific marker genes in H3K27ac 3D clusters.
a, Z-projections of H3K27ac 3D clusters identified by DBSCAN in actual data (left) and shuffled control (right), as in Fig. 3g. b, Pair auto-correlation function G(r) for H3K27ac particles and a shuffled control generated by permuting genomic positions within each cell (n = 2,736 cells). Error bars represent s.e.m. across cells. c, Heatmap showing the expression of cell-type-specific marker genes (top 200 per type) and housekeeping genes (200 genes with the lowest expression variability). d, Heatmap showing the enrichment of cell-type-specific marker genes and housekeeping genes in H3K27ac 3D clusters. e, Box plots of odds ratio quantifying the enrichment for cell type-specific marker genes compared to housekeeping genes within the H3K27ac 3D clusters (two-sided Wilcoxon signed-rank test, n = 16 cell types). Boxes indicate the median and 25th-75th percentiles, with whiskers extending to 1.5× the interquartile range. Points beyond the whiskers represent outliers.
Extended Data Fig. 7 Metacell-based correlation analysis links epigenomic features to gene expression.
a, Virtual 4 C profile showing the fraction of cells in which the Gad2 promoter is within 2 particle radii of each bin. b, Schematic of the analytical framework. Single cells were aggregated into metacells based on transcriptomic similarity. Correlations were calculated between the expression of a target gene and the signal from surrounding genomic bins for each modality. c, Venn diagram showing the overlap of significantly correlated gene-bin pairs identified by each modality. d-f, Histograms of correlation coefficients for all gene-bin pairs based on H3K27ac (d), accessibility (e), and chromatin interactions (f). Significantly correlated pairs (FDR < 0.05) are colored by the direction of correlation.
Extended Data Fig. 8 Characterization of the single-cell gene expression prediction model based on CHARM multi-omics data.
a, Distribution of Pearson’s correlation coefficients between predicted and observed gene expression on the test dataset, compared to a shuffled control. b, Comparison of Spearman correlation (ρ) of gene expression prediction between CHARM and SCARlink, computed on the 969 genes that passed method-specific QC in both pipelines. c-e, Box plots showing model performance according to gene expression level (c), gene length (d), and expression variability (e). f, Precision-recall curve evaluating E–P linkage predictions using Shapley values, compared with a correlation-based method, the ABC model and ENCODE rE2G on a gold standard set of 180 validated E–P pairs. Area under the Precision-recall curve (AUPRC) values are indicated. g, Bar plot showing the number of significant CRE-gene linkages identified in each cell type. h, Venn diagram showing the number of predicted enhancer–gene pairs identified using different combinations of modalities. i, Histogram of the genomic distance between CREs and their linked genes. The dashed line marks 200 kb. j, Histogram showing the number of TSS bypassed by distal intergenic enhancers to reach their target gene. The dashed line marks the threshold for enhancers that bypass more than one TSS. k-m, Box plots showing the maximum Shapley values across cell types for chromatin accessibility (k), H3K27ac (l) and chromatin interaction strength (m), stratified by genomic context (within gene vs. intergenic) and binned by linear distance to the TSS. In panels c-e and k-m, boxes indicate the median and 25th-75th percentiles, with whiskers extending to 1.5× the interquartile range.
Extended Data Fig. 9 Subtype-specific regulation of Satb2 by a distal enhancer associated with human intelligence traits.
a, Heatmap of Shapley values across the Satb2 locus in different cortical cell types (bottom), aligned with Satb2 expression levels (top). Excitatory neurons are grouped by expression level into Ex group 1 and Ex group 2. b, Multi-modal view of the Satb2 locus comparing Ex group 1 and Ex group 2. Tracks from top to bottom represent: mean 3D distance from the Satb2 promoter to surrounding bins (in particle radii), chromatin accessibility, and H3K27ac profiles plotted as mean fragment counts across cells in inhibitory and excitatory neurons. The heatmap displays 3D proximity matrix for the same region, indicating the percentage of cells with inter-bin distances <2 particle radii. White arrows indicate a subtype-specific chromatin loop with increased 3D proximity and elevated H3K27ac signal at the 730 kb enhancer. c and d, Genome browser views of the 1085 kb (c) and the 730 kb (d) regulatory regions, with human GWAS SNPs associated with intelligence shown in the top track.
Extended Data Fig. 10 Differential TF activity analysis at distal Satb2 and Gad2 enhancers.
a, Volcano plot showing differential TF motif activity between different groups of excitatory neuron subtypes. b, Volcano plot showing differential TF motif activity between different groups of inhibitory neuron subtypes. c, Z-score normalized ChromVAR motif activity for EGR1-3 across excitatory neuron subtypes. d, RNA expression for Egr1-3 across excitatory neuron subtypes same as in a. e, Similar to c but for NFI family across inhibitory neuron subtypes. f, Similar to d but for NFI family RNA across inhibitory neuron subtypes. g, JASPAR motif logo for the EGR and NFI family proteins. h, ChromBPNet counts-contribution tracks for the Satb2 distal enhancer (+730 kb); tracks shown for excitatory subtypes, where higher positive values indicate bases predicted to increase total accessibility in that subtype. i, Similar to h but for the Gad2 distal enhancer (−520 kb) in inhibitory subtypes.
Supplementary information
Supplementary Information (download PDF )
Supplementary Tables 1–5 and Supplementary Figs. 1–7.
Supplementary Tables (download XLSX )
Supplementary Tables 6 and 7.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Chen, Y., Liu, Z., Xu, H. et al. Gene regulatory landscape dissected by single-cell four-omics sequencing. Nature (2026). https://doi.org/10.1038/s41586-026-10322-z
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41586-026-10322-z
