Abstract
The three-dimensional (3D) organization of cis-regulatory elements (CREs) is critical in transcription control. However, capturing transcriptome, epigenome and 3D genome from the same single cells remains challenging. Here we present scHiCAR (single-cell Hi-C with assay for transposase-accessible chromatin and RNA sequencing), a plate-based combinatorial barcoding method that simultaneously profiles mRNA, open chromatin and chromosome conformation capture from the same cells. Compared to existing single-cell 3D genome methods, scHiCAR more efficiently enriches long-range cis-interactions anchored at candidate CREs (cCREs). Applied to 1.62 million mouse brain cells and complemented with a deep-learning-based loop caller, scHiCAR accurately defines cell-type-specific transcriptomes, accessible cCREs and 5-kb-resolution enhancer–promoter pairs across 22 brain cell types. scHiCAR also performs robustly in challenging tissues such as skeletal muscle, enabling trimodal single-cell-level analysis of gene regulation dynamics during muscle stem cell regeneration. By providing a scalable and cost-effective system for single-cell trimodal analysis of gene-regulatory landscapes in complex tissues, scHiCAR reveals gene-locus-specific regulatory roles of 3D genome reorganization in transcriptional control.
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 12 print issues and online access
$259.00 per year
only $21.58 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
All sequencing data were deposited to the National Center for Biotechnology Information Gene Expression Omnibus under accession number GSE305889. All chromatin loops identified by scDeepLUCIA across 22 mouse brain cell types are available from Zenodo61 or http://junglab.kaist.ac.kr/Dataset/scDeepLUCIA. The public datasets used in this study are listed in Supplementary Table 9. Source data are provided with this paper.
Code availability
Codes for scHiCAR processing and scDeepLUCIA were deposited to a GitHub repository (https://github.com/DiaoLab/scHiCAR and https://github.com/kaistcbfg/scDeepLUCIA).
References
Bonev, B. et al. Multiscale 3D genome rewiring during mouse neural development. Cell 171, 557–572 (2017).
Freire-Pritchett, P. et al. Global reorganisation of cis-regulatory units upon lineage commitment of human embryonic stem cells. Elife 6, e21926 (2017).
Greenwald, W. W. et al. Subtle changes in chromatin loop contact propensity are associated with differential gene regulation and expression. Nat. Commun. 10, 1054 (2019).
Zheng, H. & Xie, W. The role of 3D genome organization in development and cell differentiation. Nat. Rev. Mol. Cell Biol. 20, 535–550 (2019).
Kraft, K. et al. Polycomb-mediated genome architecture enables long-range spreading of H3K27 methylation. Proc. Natl Acad. Sci. USA 119, e2201883119 (2022).
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).
Lieberman-Aiden, E. et al. Comprehensive mapping of long-range interactions reveals folding principles of the human genome. Science 326, 289–293 (2009).
Nora, E. P. et al. Spatial partitioning of the regulatory landscape of the X-inactivation centre. Nature 485, 381–385 (2012).
Dixon, J. R. et al. Topological domains in mammalian genomes identified by analysis of chromatin interactions. Nature 485, 376–380 (2012).
Rao, S. S. P. et al. A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping. Cell 159, 1665–1680 (2014).
Dryden, N. H. et al. Unbiased analysis of potential targets of breast cancer susceptibility loci by Capture Hi-C. Genome Res. 24, 1854–1868 (2014).
Mumbach, M. R. et al. HiChIP: efficient and sensitive analysis of protein-directed genome architecture. Nat. Methods 13, 919–922 (2016).
Fang, R. et al. Mapping of long-range chromatin interactions by proximity ligation-assisted ChIP-seq. Cell Res. 26, 1345–1348 (2016).
Fullwood, M. J. et al. An oestrogen-receptor-α-bound human chromatin interactome. Nature 462, 58–64 (2009).
Sati, S. et al. HiCuT: an efficient and low input method to identify protein-directed chromatin interactions. PLoS Genet. 18, e1010121 (2022).
Wei, X. et al. HiCAR is a robust and sensitive method to analyze open-chromatin-associated genome organization. Mol. Cell 82, 1225–1238 (2022).
Jung, I. et al. A compendium of promoter-centered long-range chromatin interactions in the human genome. Nat. Genet. 51, 1442–1449 (2019).
Cao, J. et al. Joint profiling of chromatin accessibility and gene expression in thousands of single cells. Science 361, 1380–1385 (2018).
Ma, S. et al. Chromatin potential identified by shared single-cell profiling of RNA and chromatin. Cell 183, 1103–1116 (2020).
Zhu, C. et al. Joint profiling of histone modifications and transcriptome in single cells from mouse brain. Nat. Methods 18, 283–292 (2021).
Lee, D.-S. et al. Simultaneous profiling of 3D genome structure and DNA methylation in single human cells. Nat. Methods 16, 999–1006 (2019).
Liu, Z. et al. Linking genome structures to functions by simultaneous single-cell Hi-C and RNA-seq. Science 380, 1070–1076 (2023).
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).
Stuart, T. et al. Comprehensive integration of single-cell data. Cell 177, 1888–1902 (2019).
Mathur, R. et al. Glioblastoma evolution and heterogeneity from a 3D whole-tumor perspective. Cell 187, 446–463.e16 (2024).
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).
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).
Ramsköld, D. et al. Full-length mRNA-seq from single-cell levels of RNA and individual circulating tumor cells. Nat. Biotechnol. 30, 777–782 (2012).
Wei, X., Tran, D. & Diao, Y. HiCAR: analysis of open chromatin associated long-range chromatin interaction using low-input materials. Curr. Protoc. 3, e899 (2023).
Yang, T. et al. HiCRep: assessing the reproducibility of Hi-C data using a stratum-adjusted correlation coefficient. Genome Res. 27, 1939–1949 (2017).
Yu, M. et al. Integrative analysis of the 3D genome and epigenome in mouse embryonic tissues. Nat. Struct. Mol. Biol. 32, 479–490 (2025).
Xie, Y. et al. Droplet-based single-cell joint profiling of histone modifications and transcriptomes. Nat. Struct. Mol. Biol. 30, 1428–1433 (2023).
Zhang, R., Zhou, T. & Ma, J. Ultrafast and interpretable single-cell 3D genome analysis with Fast-Higashi. Cell Syst. 13, 798–807.e6 (2022).
Santoni, G. et al. Chromatin plasticity predetermines neuronal eligibility for memory trace formation. Science 385, eadg9982 (2024).
Gorkin, D. U. et al. An atlas of dynamic chromatin landscapes in mouse fetal development. Nature 583, 744–751 (2020).
Liu, H. et al. Single-cell DNA methylome and 3D multi-omic atlas of the adult mouse brain. Nature 624, 366–377 (2023).
Li, Y. E. et al. An atlas of gene regulatory elements in adult mouse cerebrum. Nature 598, 129–136 (2021).
Li, Y. E. et al. A comparative atlas of single-cell chromatin accessibility in the human brain. Science 382, eadf7044 (2023).
Kim, K. & Jung, I. covNorm: An R package for coverage based normalization of Hi-C and capture Hi-C data. Comput. Struct. Biotechnol. J. 19, 3149–3159 (2021).
Salameh, T. J. et al. A supervised learning framework for chromatin loop detection in genome-wide contact maps. Nat. Commun. 11, 3428 (2020).
Wolff, J. et al. Galaxy HiCExplorer 3: a web server for reproducible Hi-C, capture Hi-C and single-cell Hi-C data analysis, quality control and visualization. Nucleic Acids Res. 48, W177–W184 (2020).
Juric, I. et al. MAPS: model-based analysis of long-range chromatin interactions from PLAC-seq and HiChIP experiments. PLoS Comput. Biol. 15, e1006982 (2019).
Yang, D., Chung, T. & Kim, D. DeepLUCIA: predicting tissue-specific chromatin loops using deep learning-based universal chromatin interaction annotator. Bioinformatics 38, 3501–3512 (2022).
Fudenberg, G. et al. Formation of chromosomal domains by loop extrusion. Cell Rep. 15, 2038–2049 (2016).
Nichols, M. H. & Corces, V. G. A CTCF code for 3D genome architecture. Cell 162, 703–705 (2015).
Sanborn, A. L. et al. Chromatin extrusion explains key features of loop and domain formation in wild-type and engineered genomes. Proc. Natl Acad. Sci. USA 112, E6456–E6465 (2015).
Davidson, I. F. et al. CTCF is a DNA-tension-dependent barrier to cohesin-mediated loop extrusion. Nature 616, 822–827 (2023).
Rahme, G. J. et al. Modeling epigenetic lesions that cause gliomas. Cell 186, 3674–3685 (2023).
Schreiber, J., Singh, R., Bilmes, J. & Noble, W. S. A pitfall for machine learning methods aiming to predict across cell types. Genome Biol. 21, 282 (2020).
Visel, A., Minovitsky, S., Dubchak, I. & Pennacchio, L. A. VISTA enhancer browser—a database of tissue-specific human enhancers. Nucleic Acids Res. 35, D88–D92 (2007).
Song, M. et al. Cell-type-specific 3D epigenomes in the developing human cortex. Nature 587, 644–649 (2020).
Schmitt, A. D. et al. A compendium of chromatin contact maps reveals spatially active regions in the human genome. Cell Rep. 17, 2042–2059 (2016).
Lumpkin, E. A. et al. Math1-driven GFP expression in the developing nervous system of transgenic mice. Gene Expr. Patterns 3, 389–395 (2003).
Abdul-Aziz, D., Hathiramani, N., Phung, L., Sykopetrites, V. & Edge, A. S. B. HIC1 represses Atoh1 transcription and hair cell differentiation in the cochlea. Stem Cell Rep. 16, 797–809 (2021).
Lipton, S. A. et al. Autistic phenotype from MEF2C knockout cells. Science 323, 208 (2009).
Li, H. et al. Transcription factor MEF2C influences neural stem/progenitor cell differentiation and maturation in vivo. Proc. Natl Acad. Sci. USA 105, 9397–9402 (2008).
Yu, M. et al. SnapHiC: a computational pipeline to identify chromatin loops from single-cell Hi-C data. Nat. Methods 18, 1056–1059 (2021).
Li, X. et al. SnapHiC2: a computationally efficient loop caller for single cell Hi-C data. Comput. Struct. Biotechnol. J. 20, 2778–2783 (2022).
Zemke, N. R. et al. Conserved and divergent gene regulatory programs of the mammalian neocortex. Nature 624, 390–402 (2023).
Jung, I. & Diao, Y. Mouse frontal cortex scHiCAR 22 cell types scDeepLUCIA loops. Zenodo https://doi.org/10.5281/zenodo.18196030 (2026).
Acknowledgements
This study was supported by research funds from the Duke Regeneration Center, Duke Whitehead Scholarship, Glenn Foundation for Medical Research, American Foundation for Aging Research grants for junior faculty, NIH grants U01HL156064 and R35HG011328 to Y.D. and NIH grant 7RF1MH124612 to J.Z.H. and Y.D. Y. Xu is supported by the Center for Advanced Genomic Technologies postdoctoral fellowship and the Duke Regeneration Center Fellowship to Accelerate Career Independence. D.Y., K.K. and I.J. are supported by the Ministry of Science and ICT through the National Research Foundation of the Republic of Korea (RS-2023-00222784, 2023R1A2C3002773, RS-2023-00262527, 2022R1A5A1026413 and RS-2025-02214225), the Suh Kyungbae Foundation Fellowship and the Samsung Research Funding and Incubation Center for Future Technology of Samsung Electronics (project no. SSTF-BA2201-07). We acknowledge C. Eroglu at Duke for helping with fluorescence-activated cell sorting of astrocytes from mouse brain and Y. Zhao at Admera Health for his support in optimizing the deep sequencing strategy for scHiCAR libraries.
Author information
Authors and Affiliations
Contributions
X. Wei, Y. Xu and Y.D. conceptualized the study. X. Wei designed the scHiCAR experimental protocol and conducted all experiments, with assistance from L.Y., W. Luo, X.L., A.B.W., S.S., C.T. and Z.J.H. Bioinformatics analysis was led by Y. Xu with contributions from X. Wei, D.Y., K.K., Y. Xiang, X. Wang, K.Z., W. Li, Y.E.L., F.Y. and I.J. D.Y., K.K. and I.J. developed and validated scDeepLUCIA for scHiCAR loop calling, with input from X. Wei, Y. Xu, K.K. and Y.D. X. Wei, Y. Xu, D.Y., K.K., I.J. and Y.D. wrote the manuscript, with all coauthors providing input.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Biotechnology thanks Goncalo Castelo-Branco and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data
Extended Data Fig. 1 scHiCAR simultaneously profiles transcriptome, chromatin accessibility, and 3D genome structure in mixed human and mouse cell lines.
a. Estimated barcode complexity from four rounds of combinatorial indexing in scHiCAR. b. Species-mixing plot showing RNA read alignment to human and mouse genomes for each cell. c. Violin plots of cell type–specific marker gene expression. d–e. UMAP plots of H1 hESC and GM12878 cells based on (d) open chromatin (R2) reads and (e) chromatin contact profiles. Cells are colored by known cell identity; percentages indicate classification accuracy. f. kBET acceptance rates for RNA profiles. Box plots indicate the median, 25th–75th percentiles, and 1.5× IQR whiskers. g. LISI scores for clusters in RNA-based UMAP. h. kBET acceptance rates for chromatin accessibility (ATAC) profiles. Box plots indicate the median, 25th–75th percentiles, and 1.5× IQR whiskers. i. LISI scores for clusters in ATAC-based UMAP. j. Spearman correlation of contact matrices between replicates based on HiCRep, assessing 3D genome reproducibility. k. Venn diagrams showing overlap of scHiCAR vs. bulk HiCAR open chromatin peaks in H1 hESC (left) and GM12878 (right). Significance determined by empirical P < 1e–6. l. Genome browser tracks comparing scHiCAR and bulk HiCAR accessibility signals in GM12878 (top) and H1 hESC (bottom). m. Overlap of open chromatin peaks between two ATAC-seq replicates in H1 hESCs from the 4DN Data Portal. n. Distribution of trans, short-range (<20 kb) cis-, and long-range (≥20 kb) cis- contacts that overlapped with ATAC-seq peaks from ENCODE (ENCFF654BVB) in scHiCAR, LiMCA, and Dip-C from GM12878. o. Heatmap of pairwise Spearman correlations showing similarities in 3D genome structure among Hi-C, bulk HiCAR, and scHiCAR in the same cell lines. p. Scatterplots comparing eigenvector (left) and insulation scores (right) between bulk HiCAR and scHiCAR for H1 hESC and GM12878. q. Comparison of eigenvector and insulation scores from all vs. 10% of scHiCAR cells in H1 hESC and GM12878 datasets. r. Venn diagrams showing overlap of Peakachu-called loops between scHiCAR and bulk HiCAR at 10 kb (top) and 5 kb (bottom) in H1 hESC and GM12878. Empirical P < 1e–6. s. Venn diagrams showing overlap of HICCUPS loops between two replicates of in situ Hi-C (top, H1 hESC) and H3K4me3 PLAC-seq (bottom, K562) from 4DN Data Portal. t. Scatterplots comparing eigenvector (top) and insulation scores (bottom) for scHiCAR vs. Hi-C (left), scHiCAR vs. Dip-C (middle), and Hi-C vs. Dip-C (right) in GM12878. PCC indicates Pearson correlation.
Extended Data Fig. 2 scHiCAR enables cell type clustering based on transcriptome, chromatin accessibility, or 3D genome in mouse frontal cortex.
a. UMAP visualizations of 5,313 mouse frontal cortex cells, colored by gene activity scores determined by chromatin accessibility of representative marker gene loci. Shown are markers distinguishing neurons from non-neuronal cells (left), glutamatergic neurons (middle), and GABAergic neurons (right). b. UMAP plots highlighting gene activity scores for subtype markers of glutamatergic neurons, GABAergic neurons, and non-neuronal cells. c. Comparison of the detected gene number per cell of RNA library across different methods. The cell number (n) for each method is: 10x (n = 4,439; 4,424; 6,912), GAGE-seq (n = 3,296), HiRES (n = 399), scHiCAR (n = 9,359), and ChAIR (n = 66,606). Box plots indicate the median, 25th–75th percentiles, and 1.5× IQR whiskers.
Extended Data Fig. 3 scHiCAR accurately resolves transcriptome, chromatin accessibility, and 3D genome profiles in 22 mouse brain cell types.
a. UMAP embedding showing subclusters of L4/5 IT neurons based on scHiCAR transcriptomes. b. Heatmap of marker gene expression across L4/5 IT neuron subclusters.c. UMAP embedding showing astrocyte subclusters based on scHiCAR transcriptomes. d. Heatmap of marker gene expression across astrocyte subclusters. e. Heatmap of chromatin accessibility across 260,781 unique cCREs in 22 brain cell types. f. Boxplot of Spearman correlation coefficients (SCCs) between scHiCAR R2 signals and BICCN snATAC-seq signals at 251,496 snATAC-seq cCREs across 17 shared brain cell types. Random genomic regions (n = 251,496) were used as controls. P-value from one-sided unpaired Wilcoxon test. g. Venn diagram showing overlap of open chromatin peaks defined by R2 reads between scHiCAR (red) and bulk HiCAR (blue) in astrocytes (empirical P < 1e–6). h. Genome browser tracks comparing A/B compartments, contact matrices (5 kb bins), chromatin loops (5 kb), and open chromatin (R2 reads) signals between scHiCAR and bulk HiCAR in astrocytes. i. Venn diagrams showing overlap of Peakachu-detected chromatin interactions between scHiCAR and bulk HiCAR in Pvalb neurons (10 kb) and astrocytes (5 kb and 10 kb) (empirical P < 1e–6). j. Scatterplots comparing insulation scores between scHiCAR and bulk HiCAR in Pvalb neurons (left) and astrocytes (right). Pearson correlation coefficients (PCCs) are indicated.
Extended Data Fig. 4 Validation and performance evaluation of scDeepLUCIA for scHiCAR loop calling.
a. Relative number of loops identified by the indicated loop callers from GM12878 scHiCAR dataset (148M) before and after 10k cell downsampling (15M). b. Number of loops identified by scDeepLUCIA and Peakachu across 22 mouse brain cell clusters, ordered by the number of captured cells in each cluster. c. Receiver operating characteristic (ROC) curve and AUROC of the scDeepLUCIA model trained on the loops from downsampled astrocyte bulk HiCAR. d. Number of loops identified by scDeepLUCIA and Peakachu in downsampled L23IT-1 cell cluster. e-f. Chromatin contact maps and identified loops before and after downsampling in oligodendrocyte (e) and Pvalb (f). g. Loop recovery rate after downsampling relative to the number of loops identified in the full dataset (120,000 cells). h. Aggregate peak analysis (APA) plot of cell type specific loops identified by scDeepLUCIA. Loop anchors are centered, with ±20 neighboring bins (5kb resolution) shown. i. Genome browser snapshots displaying scDeepLUCIA loops, open chromatin peaks, compartment A/B, and gene expression for cell type specific marker gene, Car2 in oligodendrocyte (Oligo).
Extended Data Fig. 5 Design and benchmarking of scDeepLUCIA for scHiCAR loop calling.
a. Bar plot showing the number of open chromatin (R2) peaks overlapping with the top 50,000 chromatin loops identified by the indicated loop callers from the GM12878 scHiCAR dataset that are high-depth (148M reads) and downsampled (15M). b. Bar plot showing the number of loops with CTCF peaks at both anchors (left) and the number with convergent CTCF motifs (right) in the top 50,000 loops called by each method. If fewer than 50,000 loops were detected, all available loops were used. c. Receiver operating characteristic (ROC) curves comparing scDeepLUCIA (blue), Peakachu (green), and CovNorm (magenta) performance on loops identified in a downsampled (30M) mouse astrocyte bulk HiCAR dataset. Methods unable to detect loops (HiCCUPS, HiCExplorer, MAPS) are represented with AUROC = 0. d. Left: Bar plots showing the percentage of 5kb loops called by scDeepLUCIA and Peakachu that are also recovered at 10kb and 50kb resolution (Peakachu does not support 50kb) from H1-hESC (226M) and GM12878 (227M) bulk HiCAR data. Middle and right: The fraction of multi-resolution loops overlapping with open chromatin peaks (middle) or CTCF peaks with motif orientation (right) in GM12878. Input sequencing depth was adjusted to match each method’s optimal range. e. ROC curves and AUROC scores for scDeepLUCIA models trained on either mouse (7 brain cell types) or human (5 cell lines) bulk HiCAR datasets, tested on downsampled mouse astrocyte (30M) and human HepG2 (30M) bulk HiCAR datasets. f. ROC curve and AUROC for a scDeepLUCIA model trained on mouse brain scHiCAR data (6 most abundant cell types: L23IT-1, L45IT, L6CT, L6b, Oligo, and Astrocyte) tested on the human HepG2 bulk HiCAR dataset. g. ROC curve and AUROC of the scDeepLUCIA model (trained on mouse bulk HiCAR) evaluated on downsampled GM12878 H3K27ac HiChIP data (3.91M). h. Venn diagram showing the overlap between scDeepLUCIA and Peakachu-identified loops from full-depth GM12878 H3K27ac HiChIP data. Overlap significance was assessed by empirical P-value (P < 2.2e-6).
Extended Data Fig. 6 Identification of enhancer–gene pairs in mouse brain cells using scHiCAR data.
a. Proportion of CRE–gene pairs that cross topologically associating domain (TAD) boundaries. Comparisons were made between pairs predicted by BICCN snATAC-seq based on co-accessibility (light blue) and pairs identified by scHiCAR chromatin interactions (dark blue) across different brain cell types. For fair comparison, the number of co-accessible pairs (n = 418,528) was comparable to the number of scHiCAR pairs (n = 440,710) by selecting top-ranked pairs by co-accessibility score. b. Enrichment of VISTA-validated enhancers among different sets of distal cCREs. Compared were: (i) snATAC-seq distal cCREs linked to gene promoters by the ABC model (ABC score > 0.05) and (ii) snATAC-seq distal cCREs linked by scHiCAR chromatin interactions. Odds ratios were calculated using Fisher’s exact test. For fair comparison, both sets contained similar numbers of cCREs (47,332 for ABC and 43,290 for scHiCAR). c. Genome browser view showing a VISTA-validated enhancer (mm1451.0) that interacts with the Grk3 promoter in astrocytes. Chromatin interactions called by scDeepLUCIA are shown as black squares on the contact matrix. d. Log-normalized expression of genes with promoter accessibility linked to cCREs (red) versus unlinked genes (pink) across brain cell types. Box plots indicate the median, 25th–75th percentiles, and 1.5× IQR whiskers. P-values from two-sided unpaired Wilcoxon tests. The number of samples from left to right are 7,385; 5,066; 7,275; 4,990; 7,269; 5,227; 5,477; 4,212; 7,327; 4,501; 7,616; 5,143; 6,731; 5,012; 5,555; 3,756; 7,269; 4,710; 7,020; 5,135; 6,551; 5,435; 2,009; 2,448; 2,243; 1,708; 7,361; 4,609; 7,857; 4,859; 6,211; 4,109; 1,901; 1,611; 6,548; 5,473; 7,136; 5,421; 5,748; 5,960; 5,741; 5,707; 4,861; 5,419. The p-values are 1.37e-38, 2.45e-35, 6.65e-41, 3.05e-22, 1.03e-30, 7.47e-40, 1.32e-37, 2.78e-14, 1.89e-27, 1.88e-47, 2.10e-50, 1.01e-19, 0.00024, 3.78e-24, 6.45e-24, 6.14e-19, 3.23e-10, 2.21e-36, 4.91e-28, 2.85e-33, 7.86e-54, 1.04e-28. e. An example of positive association between gene expression and the number of scHiCAR chromatin interactions in L23IT-1 neurons. Shown are RNA expression (red) and chromatin accessibility (blue) tracks. f. Scatterplot showing correlation between changes in the number of scHiCAR chromatin interactions and changes in gene expression between two brain cell types. The dashed line indicates the linear regression fit. The Pearson correlation coefficient (PCC) and associated p-value were calculated using a two-sided Pearson correlation test. g. Density plot showing the distribution of Pearson correlation coefficients (PCCs) between gene expression and chromatin accessibility at distal CREs for all CRE–gene pairs. A subset of 20,270 pairs with the strongest positive correlations (FDR < 0.05) was defined as high-confidence enhancer–gene pairs (red). Randomized pairings (gray) serve as a control. h. Comparison of enhancer–gene correlation scores between scHiCAR and Paired-seq. Top 34,473 enhancer–gene pairs predicted by scHiCAR were compared with the same number of pairs from Paired-seq based on mRNA and H3K27ac signal correlation. Distribution of Spearman correlation coefficients is shown. P-value calculated by two-sided unpaired Wilcoxon test.
Extended Data Fig. 7 scHiCAR analysis of mouse skeletal muscle regeneration.
a. UMAP visualizations of mouse muscle cells colored by gene expression for representative marker genes of different cell types as indicated. b. UMAP visualizations of mouse muscle cells colored by gene activity scores (chromatin accessibility) for representative marker genes of different cell types. c. Smoothed spline regression curves of gene expression (top), gene activity (middle), and scA/B value (bottom) of Myog over pseudotime of muscle regeneration. The dashed line marks the pseudotime at which the predicted value is maximal. d. Single-cell contact matrices centered on Myog locus at three selected cells (cell 1, 2, 3) along pseudotime, as marked in Fig. 6k.
Extended Data Fig. 8 Benchmarking scHiCAR data quality against HiRES, GAGE-seq, and ChAIR.
a. Comparison of chromatin contact matrices from scHiCAR (above diagonal) versus GAGE-seq (left), HiRES (middle), and ChAIR (right) in astrocytes. Matrices from GAGE-seq, HiRES, and ChAIR are shown below the diagonal. b. Scatterplots comparing eigenvector values (top) and insulation scores (bottom) between scHiCAR and GAGE-seq (left), HiRES (middle), and ChAIR (right) in astrocytes. Pearson correlation coefficients (PCCs) are indicated. c. Loop overlap between scHiCAR loops (called by Peakachu at 5 kb resolution) and loops from HiRES and GAGE-seq (called by snapHiC2) in Ex1 and L23IT brain cell types. Overlap is highly significant (empirical p < 2.2e–16). d. Loop overlap between scHiCAR loops (called by scDeepLUCIA at 5 kb resolution) and loops from HiRES and GAGE-seq (called by snapHiC2) in Ex1 and L23IT brain cell types. Overlap is highly significant (empirical p < 2.2e–16).
Extended Data Fig. 9 Comparison of scHiCAR-unique versus snATAC-seq–unique open chromatin peaks.
a. MACS2 P-value distributions for open chromatin peaks in H1 hESCs (top) and GM12878 cells (bottom). Shown are shared peaks (scHiCAR + ATAC-seq; red) and ATAC-seq unique peaks (blue). Box plots indicate the median, 25th–75th percentiles, and 1.5× IQR whiskers. Statistical significance was assessed by the Wilcoxon rank-sum test. b. Heatmaps showing enrichment of bulk ATAC-seq and CTCF ChIP-seq signals in H1 hESCs (left) and GM12878 cells (right). Signal intensity is centered on peak regions from three categories: shared scHiCAR–ATAC peaks (top), scHiCAR-unique peaks (middle), and random genomic regions (bottom).
Supplementary information
Supplementary Information
Supplementary Methods and Legends for Supplementary Tables 1–9.
Supplementary Tables 1–9
Oligo sequences used in this study. cCREs of scHiCAR and bulk HiCAR from H1 and GM12878. Full list of cCREs in 22 mouse brain cell types identified by scHiCAR. The 10-kb chromatin interactions of scHiCAR and HiCAR called by Peakachu from Pvalb and astrocyte. The 5-kb chromatin interactions called by Peakachu in 22 mouse brain cell types. The 5-kb chromatin interactions called by scDeepLUCIA in 22 mouse brain cell types. The CRE–gene pairs identified by scHiCAR and by BICCN snATAC-seq that cross TAD boundaries. The 20,270 CRE–gene pairs with a positive correlation. List of public datasets used in this study.
Source data
Source Data Fig. 1
Statistical source data.
Source Data Fig. 2
Statistical source data.
Source Data Fig. 3
Statistical source data.
Source Data Fig. 4
Statistical source data.
Source Data Fig. 5
Statistical source data.
Source Data Fig. 6
Statistical source data.
Source Data Extended Data Fig. 1
Statistical source data.
Source Data Extended Data Fig. 2
Statistical source data.
Source Data Extended Data Fig. 3
Statistical source data.
Source Data Extended Data Fig. 4
Statistical source data.
Source Data Extended Data Fig. 5
Statistical source data.
Source Data Extended Data Fig. 6
Statistical source data.
Source Data Extended Data Fig. 9
Statistical source data.
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
Wei, X., Xu, Y., Yang, D. et al. Trimodal single-cell profiling of transcriptome, epigenome and 3D genome in complex tissues with scHiCAR. Nat Biotechnol (2026). https://doi.org/10.1038/s41587-026-03013-7
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s41587-026-03013-7
