Reconstruction of Diploid High-Order 3D Genome Interactions from Long Noisy Concatemers

Reconstruction of Diploid High-Order 3D Genome Interactions from Long Noisy Concatemers | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Technical Report Reconstruction of Diploid High-Order 3D Genome Interactions from Long Noisy Concatemers Chuan-Le Xiao, Ying Chen, Zhuo-Bin Lin, Shao-Kai Wang, Bo Wu, and 12 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3486881/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 04 Mar, 2025 Read the published version in Nature Structural & Molecular Biology → Version 1 posted You are reading this latest preprint version Abstract Differential high-order chromatin interactions between homologous chromosomes play pivotal roles in many biological processes. However, their elucidation has been hindered by technical difficulties. Traditional 3C methods mainly expose 2-way interactions and offer limited haplotype information. In response, we addressed challenges in harnessing merely nanopore high-order concatemer (Pore-C) reads to delineate diploid high-order chromatin interactions. By training a cutting-edge deep learning model and making statistical analysis, we achieved superior SNV calling and haplo-tagging for noisy short monomers. Learning the haplotype characteristics of high-order concatemers allowed us to devise a progressive haplotype imputation strategy, which elevated the haplotype informative Pore-C contact rate by 14.1-fold to 76% in the HG001 cell line, eclipsing Hi-C's rate by over ten times. Overall, our diploid 3D genome interactions surpassed conventional methods in noise reduction and contact distribution uniformity, with over one magnitude advantages in haplotype informative contact density and genomic coverage rate. Dip3D enabled the unveiling of haplotype high-order interactions in many previously overlooked genomic regions and the investigation of their relationship with allele-specific expression, especially concerning X-chromosome inactivation. Our approach stands out as a robust and accessible solution for the high-quality reconstruction of diploid high-order 3D genome interactions. Biological sciences/Computational biology and bioinformatics/High-throughput screening Biological sciences/Genetics/Genomics/Epigenomics Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Full Text Additional Declarations There is NO Competing Interest. Supplementary Files 20241024SupplementaryTables.xlsx Supplementary Tables 20241024Dip3DSupplementaryNotes.pdf 20241024SupplementaryFigures.docx Cite Share Download PDF Status: Published Journal Publication published 04 Mar, 2025 Read the published version in Nature Structural & Molecular Biology → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-3486881","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Technical Report","associatedPublications":[],"authors":[{"id":371265518,"identity":"26242cd2-f7bf-4a9d-a43e-cf77bd9efba2","order_by":0,"name":"Chuan-Le Xiao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2UlEQVRIiWNgGAWjYDACZiBmbAAS7AyMDxIqbEjRwszAbPDgTBqRNkG1sEk+bDtEWLXBceaHD37uOCxvzsx8rCKB7QADf3t3Al4tks1sxoa9Zw4b7mxmS7uRwHOHQeLM2Q14tfAzM5hJM7YdZtxwmMfsRoLEMwYDiVz8WtiY2b+BtNhvOMz/rSDB4DBhLfzMPGBbEoG2sDEkJBChRbKZp9iwty09ecNhNmOJhANpPAT9YnD++MYHP9usbTccb3748ec/Gzn+9l78WjAAD2nKR8EoGAWjYBRgBQDnbUb+GaTUTAAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0002-4680-0682","institution":"Sun Yat-sen University","correspondingAuthor":true,"prefix":"","firstName":"Chuan-Le","middleName":"","lastName":"Xiao","suffix":""},{"id":371265519,"identity":"252b68fa-7424-441f-a327-d5700270400f","order_by":1,"name":"Ying Chen","email":"","orcid":"","institution":"Sun Yat-sen University","correspondingAuthor":false,"prefix":"","firstName":"Ying","middleName":"","lastName":"Chen","suffix":""},{"id":371265520,"identity":"35ce67c0-733f-4a0e-beda-edbe8a80000b","order_by":2,"name":"Zhuo-Bin Lin","email":"","orcid":"https://orcid.org/0000-0002-6076-5879","institution":"Sun Yat-sen University","correspondingAuthor":false,"prefix":"","firstName":"Zhuo-Bin","middleName":"","lastName":"Lin","suffix":""},{"id":371265521,"identity":"ac7682c0-b97a-4cff-aa5a-fc135763dc5b","order_by":3,"name":"Shao-Kai Wang","email":"","orcid":"","institution":"David R. Cheriton School of Computer Science, University of Waterloo, Ontario, Canada;","correspondingAuthor":false,"prefix":"","firstName":"Shao-Kai","middleName":"","lastName":"Wang","suffix":""},{"id":371265522,"identity":"5798e937-fda0-4c3d-ab2f-1f9facd57d87","order_by":4,"name":"Bo Wu","email":"","orcid":"","institution":"Sun Yat-sen University","correspondingAuthor":false,"prefix":"","firstName":"Bo","middleName":"","lastName":"Wu","suffix":""},{"id":371265523,"identity":"08d3116d-103b-43be-a170-46fe88684eda","order_by":5,"name":"Longjian Niu","email":"","orcid":"","institution":"Shenzhen Eye Hospital","correspondingAuthor":false,"prefix":"","firstName":"Longjian","middleName":"","lastName":"Niu","suffix":""},{"id":371265524,"identity":"d5b5b3f7-ceef-4bfe-a519-749235ea9f5a","order_by":6,"name":"Jiayong Zhong","email":"","orcid":"https://orcid.org/0000-0001-9357-1117","institution":"Sun Yat-sen University","correspondingAuthor":false,"prefix":"","firstName":"Jiayong","middleName":"","lastName":"Zhong","suffix":""},{"id":371265525,"identity":"49c6b483-ff5f-4546-a884-d971783a605c","order_by":7,"name":"Yi-Meng Sun","email":"","orcid":"","institution":"Sun Yat-sen University","correspondingAuthor":false,"prefix":"","firstName":"Yi-Meng","middleName":"","lastName":"Sun","suffix":""},{"id":371265526,"identity":"b4fb9d19-6c9d-4a8e-905c-a8e6e36f65b6","order_by":8,"name":"Zhenxian Zheng","email":"","orcid":"https://orcid.org/0000-0002-6546-2324","institution":"The University of Hong Kong","correspondingAuthor":false,"prefix":"","firstName":"Zhenxian","middleName":"","lastName":"Zheng","suffix":""},{"id":371265527,"identity":"802a605e-bb49-4f93-b8ae-8260173c4769","order_by":9,"name":"Xin Bai","email":"","orcid":"","institution":"Sun Yat-sen University","correspondingAuthor":false,"prefix":"","firstName":"Xin","middleName":"","lastName":"Bai","suffix":""},{"id":371265528,"identity":"1bbeb447-4424-42cd-9ec1-b784fa82d602","order_by":10,"name":"Luo-Ran Liu","email":"","orcid":"","institution":"Sun Yat-sen University","correspondingAuthor":false,"prefix":"","firstName":"Luo-Ran","middleName":"","lastName":"Liu","suffix":""},{"id":371265529,"identity":"8fa8c6ce-4018-40eb-88d1-9701a3b309d1","order_by":11,"name":"Wei Xie","email":"","orcid":"https://orcid.org/0000-0003-2126-3849","institution":"Tsinghua University","correspondingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"Xie","suffix":""},{"id":371265530,"identity":"483be139-d532-4840-9021-50e671d08deb","order_by":12,"name":"Wei chi","email":"","orcid":"","institution":"Shenzhen Eye Hospital","correspondingAuthor":false,"prefix":"","firstName":"Wei","middleName":"","lastName":"chi","suffix":""},{"id":371265531,"identity":"bb199b56-309a-4a77-81bf-95c282278a5f","order_by":13,"name":"Ruibang Luo","email":"","orcid":"","institution":"The University of Hong Kong","correspondingAuthor":false,"prefix":"","firstName":"Ruibang","middleName":"","lastName":"Luo","suffix":""},{"id":371265532,"identity":"bce0bd85-8eb3-4ade-86ba-8bc2cd06e4f2","order_by":14,"name":"Chunhui Hou","email":"","orcid":"https://orcid.org/0000-0002-8339-1857","institution":"Kunming Institute of Zoology, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Chunhui","middleName":"","lastName":"Hou","suffix":""},{"id":371265533,"identity":"c8203561-bd50-4c12-992d-16ec599d7018","order_by":15,"name":"Feng Luo","email":"","orcid":"https://orcid.org/0000-0002-4813-2403","institution":"Clemson University","correspondingAuthor":false,"prefix":"","firstName":"Feng","middleName":"","lastName":"Luo","suffix":""},{"id":371265534,"identity":"d7141d73-8e40-4e60-8222-5a124279de56","order_by":16,"name":"Tiantian Ye","email":"","orcid":"","institution":"Xianghu Laboratory","correspondingAuthor":false,"prefix":"","firstName":"Tiantian","middleName":"","lastName":"Ye","suffix":""}],"badges":[],"createdAt":"2023-10-24 16:26:32","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3486881/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3486881/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41594-025-01512-w","type":"published","date":"2025-03-04T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":67806826,"identity":"b11dbdca-6911-4d94-af8a-1d8bbc7a0a66","added_by":"auto","created_at":"2024-10-30 01:40:31","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1205768,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe complete procedure of constructing diploid high-order 3D genome interactions from high-order concatemers. \u003c/strong\u003e(1) Pore-C concatemers are decomposed into monomers that are aligned to the reference genome using Falign. Falign was specifically designed for mapping high-order concatemers, which utilizes restriction enzyme sites to precisely identify monomer edges and keeps records of monomer orders\u003csup\u003e24\u003c/sup\u003e. The order of the monomers in concatemers is indicated by different colors in this Figure. (2) Genome-wide SNVs are called from aligned Pore-C reads using Pore-C-model trained using Clair3\u003csup\u003e25\u003c/sup\u003e in this study. (3) Pore-C concatemers are reformatted and used for genome-wide SNV phasing via HapCUT2. (4) Noisy Pore-C monomers are haplo-tagged using Dip3D’s haplo-tagging module that has implemented three optimized filters. Monomers are assigned to haplotype 1 (HP1, orange circles) and 2 (HP2, green circles). Un-tagged monomers are presented as gray circles. (5) Haplotype imputation is carried out for untagged monomers in high-order reads, guided by the three rules learned in this study. (6) The final output is groups of ordered h-cis monomers. Interaction paths are depicted as dashed curves linking monomers according to their order in the original concatemers, and terminal monomers of each h-cis group are marked with squares. (7) Exemplary applications of the high-order diploid 3D genome interactions. Left, diploid contact matrices; center, intricate haplotype-specific 3D structures; right, haplotype-specific high-order interactions.\u003c/p\u003e","description":"","filename":"image1.tiff.png","url":"https://assets-eu.researchsquare.com/files/rs-3486881/v1/6202cbbb5f20ca1ab138cd2c.png"},{"id":67806825,"identity":"d99aea0e-378a-42fe-bdfd-1d33d9325610","added_by":"auto","created_at":"2024-10-30 01:40:31","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1475555,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStatistics on SNV calling and haplo-tagging of Pore-C monomers. A\u003c/strong\u003e,\u003cstrong\u003e \u003c/strong\u003eComparison of SNV calling performance using different models on two different human cell lines. Different depths (horizontal axes) of Pore-C data of human cell lines HG001 and HG002 were subjected to the tests with their gold standard whole-genome SNVs from Genome in a Bottle (GIAB) used as ground truth. WGS (whole-genome sequencing) and Pore-C denote the analyzed data type. WGS-model denotes the original Clair3\u003csup\u003e25\u003c/sup\u003e model\u003cstrong\u003e \u003c/strong\u003etrained using nanopore whole-genome sequencing data, and Pore-C-model denotes the guppy4 version Pore-C model retrained in this study. \u003cstrong\u003eB-D\u003c/strong\u003e, Comparison between Hi-C and Pore-C based SNV phasing in respects of switch error rate (B), hamming error rate (C), and the proportion of phased SNVs in by the major (largest) phase blocks of chromosomes (D) in HG001 dataset. \u003cstrong\u003eE\u003c/strong\u003e, Distribution of candidate haplo-tagging monomer rate among Pore-C monomers of different lengths. \u003cstrong\u003eF\u003c/strong\u003e, Statistics on overlapping heterozygous SNV (het-SNV) count of Pore-C monomers. \u003cstrong\u003eG\u003c/strong\u003e, Statistics of haplo-tagging candidate monomers in Pore-C concatemers. \u003cstrong\u003eH\u003c/strong\u003e, The relationship between h-trans rate and interaction distance for Pore-C and Hi-C pairwise contacts. Pore-C (concatemer) pairwise contacts include both adjacent and non-adjacent monomer pairs. Haplo-tagging was carried out using WhatsHap with default settings for both the Pore-C (190×) and Hi-C (531×) data. \u003cstrong\u003eI\u003c/strong\u003e-\u003cstrong\u003eK\u003c/strong\u003e, H-trans rates, and candidate haplo-tagging monomer loss rates under different thresholds of monomer mapping identity (\u003cstrong\u003eI\u003c/strong\u003e), mapping quality (\u003cstrong\u003eJ\u003c/strong\u003e), and the length of mismatch-free (relative to the reference genome) SNV context in a monomer (\u003cstrong\u003eK\u003c/strong\u003e). Haplo-tagging candidates denote monomers overlapping with phased heterozygous SNVs. The ratios of dumped haplo-tagging candidates (loss) corresponding to different filter thresholds are drawn as orange curves in Figures I-L. \u003cstrong\u003eL\u003c/strong\u003e, Cumulative h-trans rate reduction and candidate monomer loss after applying three filters.\u003c/p\u003e","description":"","filename":"image2.tiff.png","url":"https://assets-eu.researchsquare.com/files/rs-3486881/v1/174c08117190acb2f75ffb6d.png"},{"id":67806834,"identity":"8ab52c22-2f82-4760-bcf9-1a390848301a","added_by":"auto","created_at":"2024-10-30 01:40:32","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1200718,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCharacteristics of haplotype distribution in high-order concatemers. \u003c/strong\u003eGraphs were drawn based on analysis results of 54× Pore-C data of the F1 mice. \u003cstrong\u003eA, \u003c/strong\u003eGeneration of high-heterozygosity F1 mice (C57BL/6J × PWK/PhJ). \u003cstrong\u003eB, \u003c/strong\u003eDistribution of read monomer counts and monomer haplo-tagging ratios in the analyzed Pore-C dataset. \u003cstrong\u003eC\u003c/strong\u003e, The relationship between concatemer haplo-tagged monomer count and read haplotype. Read haplotype categories: H1, reads containing maternal haplotype monomers and no paternal monomer; H2, reads containing paternal monomers and no maternal monomer; h-trans, reads containing both maternal and paternal monomers. A total of 1,133,402 reads were subjected to statistics in this graph. \u003cstrong\u003eD\u003c/strong\u003e, The relationship between interaction genomic distance and h-trans ratio of concatemer pairwise contacts. The dashed line helps identify the corresponding interaction distance (horizontal axis) with a 5% h-trans ratio on the curve. \u003cstrong\u003eE\u003c/strong\u003e, Two simplified homologous chromatin interaction modes and corresponding minority haplotype monomer distribution. The orange and green rectangles represent the majority and minority haplotype monomers, respectively. Majority haplotype monomers account for ≥1/2 of the monomers in a concatemer, while minority haplotype monomers account for ≤1/2. In cases where the counts of monomers from two different haplotypes are equal within a concatemer, the majority and minority haplotypes are assigned randomly. For the intertwining model, although being depicted with equal abundance in the diagram, the frequencies of minority haplotypes may differ between the terminal positions (1 and 3) and the center position (2). \u003cstrong\u003eF\u003c/strong\u003e, Minority haplotype monomer distributions in 3- to 6-monomer (haplo-tagged) reads of the realistic F1 mice dataset. In all panels, statistics were made on reads containing only one minority haplotype monomer. Parameters on top of each panel: \u003cem\u003eN\u003c/em\u003e, total number of concatemers subjected to statistics; \u003cem\u003eEF\u003c/em\u003e, enrichment fold of terminal minority haplotype monomers; \u003cem\u003eP\u003c/em\u003e, P value of Chi-square test against the null hypothesis that minority haplotype monomers are randomly distributed in concatemers. \u003cstrong\u003eG\u003c/strong\u003e, The relationship between h-trans (corresponding to its most adjacent haplo-tagged monomer in concatemer) rate of the terminal monomer and the number of consecutive h-cis neighboring monomers.\u003c/p\u003e","description":"","filename":"image3.tiff.png","url":"https://assets-eu.researchsquare.com/files/rs-3486881/v1/2b32e9d3b068be0679e7c479.png"},{"id":67806835,"identity":"ef81e700-3803-4a23-8ee6-f89590dc3e74","added_by":"auto","created_at":"2024-10-30 01:40:32","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":2111213,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eHaplotype imputation for high order concatemers of HG001. A\u003c/strong\u003e, Execution of distance, bridge, and dominant haplotype rules in haplotype imputation for high-order concatemers. Each of rectangles A-F represents a monomer in a six-monomer Pore-C concatemer, with monomers of different haplotypes framed in orange (hap1) and green (hap2). \u003cstrong\u003eDistance rule application\u003c/strong\u003e: Short-range contacts (arcs) under an interaction distance threshold corresponding 5% h-trans ratio are used to infer haplotypes (indicated by square colors) based on SNV-haplo-tagged (orange or green colored) monomers. Orange and green arcs denote hap1 and hap2 inferences, respectively. Only untagged monomers with a single haplotype inference or those inferred to be one haplotype with at least twice as many short-range contacts as the second haplotype are imputed. Untagged monomers with unconfident conflicting inferences are discarded. \u003cstrong\u003eBridge rule application\u003c/strong\u003e: If two monomers sharing the same haplotype were connected (termed a 'bridge') and no intervening monomers had a contrasting haplotype, the untagged monomers on the bridge were assigned the haplotype of the two monomers. The concatemers are split at h-trans junctions and untagged monomers flanked by h-trans monomers are discarded, resulting in split h-cis monomer groups. After applying the bridge haplotype rule, monomers of the same haplotype were rejoined. \u003cstrong\u003eDominant haplotype rule application\u003c/strong\u003e: Untagged monomers within each split group were then assigned to the predominant haplotype if it comprised ≥80% of the monomers, and then h-cis monomers in each concatemer were re-connected to form final haplotype-assigned h-cis groups. \u003cstrong\u003eB\u003c/strong\u003e, Diploid contact heatmaps of HG001 before and after each of the imputation steps. The gray lines denote the corresponding stages between diagram (\u003cstrong\u003eA\u003c/strong\u003e) and the diploid contact heatmaps (\u003cstrong\u003eB\u003c/strong\u003e). \u003cstrong\u003eC\u003c/strong\u003e, The correlation between interaction distance and h-trans rate in HG001 after applying our haplo-tagging module. 'Pore-C' indicates all pairwise contacts (including both adjacent and non-adjacent contacts) from Pore-C reads, and ‘Pore-C (adjacent)' refers to pairwise contacts between directly neighboring monomers in concatemers. Interaction distance thresholds for distinguishing short-range and long-range contacts are set at the 5% h-trans contact rate level. \u003cstrong\u003eD-F\u003c/strong\u003e, Proportions of haplo-tagged monomers (\u003cstrong\u003eD\u003c/strong\u003e), haplotype-assigned contacts (\u003cstrong\u003eE\u003c/strong\u003e), and haplotype-assigned concatemers (\u003cstrong\u003eF\u003c/strong\u003e) after each of the three imputation steps.\u003c/p\u003e","description":"","filename":"image4.tiff.png","url":"https://assets-eu.researchsquare.com/files/rs-3486881/v1/4557c9a08419a577421daa2a.png"},{"id":67807063,"identity":"a06b141e-c4a8-47b7-b745-c626ffe80794","added_by":"auto","created_at":"2024-10-30 01:48:31","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":7862944,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eValidation of canonical diploid 3D genome structures in HG001. A \u003c/strong\u003eand\u003cstrong\u003eB\u003c/strong\u003e,\u003cstrong\u003e \u003c/strong\u003ecorresponding\u003cstrong\u003e \u003c/strong\u003ecorrelation coefficients of A/B compartment eigenvector scores (A) and TAD insulation scores (B) between homologous chromosomes in Pore-C (Dip3D) and Hi-C diploid contact matrices. Framed dots in the top right corner of graph A are zoomed out (indicated by dashed lines) for better visualization in a new coordinate system. \u003cstrong\u003eC\u003c/strong\u003e, Dip3D diploid contact matrices on an exemplary autosome (chr2). Color scale (from white to red): Zero contact frequency is represented with white color, and the left bottom red square indicates the color of the maximum haplotype-assigned contact count (number right to the square) in the panel, which also applies to the following panels. \u003cstrong\u003eD\u003c/strong\u003e-\u003cstrong\u003eF\u003c/strong\u003e, Hi-C and Pore-C based diploid contact heatmaps of the entire chromosome X (\u003cstrong\u003eD\u003c/strong\u003e), the superdomain (at ~115Mb) region (\u003cstrong\u003eE\u003c/strong\u003e), and the superloop (formed between tandem repeats DXZ4 and FIRRE) region (\u003cstrong\u003eF\u003c/strong\u003e). The regions indicated by red arrows between the dashed horizontal lines in \u003cstrong\u003eD\u003c/strong\u003ewere zoomed out as \u003cstrong\u003eE\u003c/strong\u003e, and the circled regions in \u003cstrong\u003eE\u003c/strong\u003e were further zoomed out as \u003cstrong\u003eF\u003c/strong\u003e. \u003cstrong\u003eG\u003c/strong\u003e, Hi-C, and Pore-C diploid contact heatmap at the \u003cem\u003eH19\u003c/em\u003e/\u003cem\u003eIgf2\u003c/em\u003e Distal Anchor Domain (HIDAD).\u003c/p\u003e","description":"","filename":"image5.tiff.png","url":"https://assets-eu.researchsquare.com/files/rs-3486881/v1/2e953f1f9b179f1c43a6f148.png"},{"id":67806827,"identity":"bc02567b-98b3-4807-8f22-b14b5143d343","added_by":"auto","created_at":"2024-10-30 01:40:31","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":3900030,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEvaluation of contact distribution and 3D structures in diploid 3D genomes.\u003c/strong\u003e Panels \u003cstrong\u003eA\u003c/strong\u003e and \u003cstrong\u003eB\u003c/strong\u003e depict the results based on 531× Hi-C and 190× Pore-C data of HG001. \u003cstrong\u003eA\u003c/strong\u003e, resolution of Pore-C (Dip3D) and Hi-C based diploid contact matrices. Resolution denotes the minimum bin size that ensures an overlap of ≥1,000 contacts in at least 80% (dashed line) of bins. \u003cstrong\u003eB\u003c/strong\u003e, distribution of poorly covered ‘difficult’ regions by Dip3D and Hi-C diploid contact matrices. Regions with \u0026lt;10× haplotype-assigned monomer coverage were regarded as poorly covered regions (Supplementary Table 5). The ‘difficult’ regions were obtained from the GRCh37 genome stratification files from Genome in A Bottle (GIAB). Homopolymer, diTR, triTR, and tetraTR (simple repeats) denote tandemly repeated single-, di-, tri-, and tetra-nucleotide motifs. Low mappability regions have other homologous regions in the reference genome as identified with different stringency parameters involving read length (l), number of mismatches (m), and number of indels (e). The horizontal axis: GC content, GC ratio intervals; simple repeats and tandem repeats, the length intervals of annotated repeat regions; segmental duplications, ‘all’ annotated segmental duplicated regions, ‘\u0026gt;10kb’ segmental duplicated regions, and ‘\u0026gt;10kb-5dups’ is further subset to regions \u0026gt;10kb covered by more than 5 segmental duplications with \u0026gt;99% identity; low mappability, ‘l100_m2_e1’ (l=100, m=2, e=1) and ‘l250_m0_e0’ (l=250, m=0, e=0) denote low mappability regions identified with two different stringency parameters by GIAB, and ‘all’ is a union of them. Panels \u003cstrong\u003eC-G\u003c/strong\u003e compared the performances of equal contact-sized (160M per haplotype) Pore-C (Dip3D) and Hi-C (Rao et al., 2014) diploid contact matrices. \u003cstrong\u003eC\u003c/strong\u003e, comparison of diploid contact frequency distributions in NGS ‘difficult’ regions. The relative contact frequency was normalized by dividing the mean haplotype-assigned contact frequency (1.0 on the axes) of the non-‘difficult’ regions. For both Hi-C and Pore-C, the maternal (left boxplots) and paternal (right boxplots) distributions were drawn separately. \u003cem\u003eFc\u003c/em\u003e denotes fold change. \u003cstrong\u003eD\u003c/strong\u003e, distribution of Hi-C and Pore-C haplotype-assigned contact frequencies across whole-genome 50kb bins. The peaks of the distribution curves are denoted using dashed lines and labeled with the mode of bin contact frequencies. Std denotes standard deviation, and IQR is an abbreviation of interquartile range, which is the contact frequency difference between the 75\u003csup\u003eth\u003c/sup\u003e and 25\u003csup\u003eth\u003c/sup\u003e percentiles of the bins. \u003cstrong\u003eE\u003c/strong\u003e, QuASAR-QC scores of Pore-C and Hi-C based diploid contact matrices at different bin sizes. Evaluation of the maternal (left) and paternal (right) haplotype matrices was carried out separately. \u003cstrong\u003eF\u003c/strong\u003e, reproducibility of TADs detected from unphased Hi-C contacts by Hi-C and Pore-C-based paternal and maternal contact matrices. All TADs were called at 50 kb resolution. The 1,855 unphased TADs were identified using the same Hi-C data used for building the diploid contact matrix. TADs with a reciprocal overlap of ≥80% were regarded as the same. In genomic regions without allelically different 3D structures, unphased TADs are expected to be reproducible from both paternal and maternal contact matrices. Under ideal conditions where the Hi-C contacts are all phased, most unphased TADs are expected to be detected from the paternal/maternal contact matrices. \u003cstrong\u003eG\u003c/strong\u003e, TADs (unphased) reproduced from Hi-C and Pore-C diploid contact matrices in an exemplary region.\u003c/p\u003e","description":"","filename":"image6.tiff.png","url":"https://assets-eu.researchsquare.com/files/rs-3486881/v1/c1c36a74d17b9750a319bf7b.png"},{"id":67806831,"identity":"0f3cbf64-d2d2-4dab-b865-bbcc040928f7","added_by":"auto","created_at":"2024-10-30 01:40:31","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":2203413,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eHaplotype-specific high-order interaction at genic promoter regions. A\u003c/strong\u003e, calculation of single-chromatin distal to proximal contact ratio (SDPR). An exemplary concatemer containing five haplotype-assigned monomers is illustrated at the top. The target monomer (red bar in the graph) overlaps with a genomic region of interest. Monomers within 1.5 Mb distance from the target monomer are considered proximal monomers (green), while those outside this range are considered distal monomers (black). For a target monomer in a concatemer, SDPR is calculated by dividing the count of its proximal monomers by the count of its distal monomers. \u003cstrong\u003eB\u003c/strong\u003e, calculation of multi-enhancer interacting concatemer ratio (MER). NE, SE, and ME denote concatemers in which a target promoter (red bar) overlapping monomer (in red shadow) interacts with none, a single, and more than one enhancer (orange bars) overlapping monomers, respectively. For a given promoter region, MER is calculated as the ratio of ME concatemers to the total number of NE, SE, and ME-type concatemers for that promoter. NER (none-enhancer interacting ratio) and SER (single-enhancer interacting ratio) shown in graph F are calculated similarly to MER. The promoter regions in the analyses were obtained from the GRCh38 GENCODE v29 gene annotation\u003csup\u003e54\u003c/sup\u003e, and the enhancers were derived from the ENCODE cRES v3 functional element annotation database\u003csup\u003e55\u003c/sup\u003e. \u003cstrong\u003eC \u003c/strong\u003eand\u003cstrong\u003e D, \u003c/strong\u003edistribution of allelic\u003cstrong\u003e \u003c/strong\u003esingle chromatin distal to proximal contact ratios (SDPRs) and multi-enhancer interacting ratios (MERs)\u003cstrong\u003e \u003c/strong\u003eat the promoter regions of autosomal genes and X chromosomal genes. Autosomal genes were classified as genes with maternal/paternal allele-specific expression (ASE) according to previous reports\u003csup\u003e15,40\u003c/sup\u003e and other genes that have not been reported with ASE. The ASE genes had at least 2-fold expression quantification difference between alleles tested with \u0026lt;0.05 adjusted-p values. X chromosomal genes were classified as genes subjected to X chromosome inactivation (XCI) and those escaped from XCI based on the study of Balaton et al., (2015).\u003cstrong\u003e E\u003c/strong\u003e, SDPR and monomer distribution of concatemers overlapping the promoters of a gene subjected to XCI\u003cem\u003e \u003c/em\u003e(\u003cem\u003eRBM3\u003c/em\u003e, left) and another gene escaped from XCI (\u003cem\u003eIQSEC2\u003c/em\u003e, right). In the top h1 and h2 panels, each row represents a concatemer. Proximal (green dots) and distal (black dots) monomers of the target promoters are differentially colored. The distribution of monomer abundance in the genomic intervals is shown as histograms in the bottom panel. \u003cstrong\u003eF\u003c/strong\u003e, distribution of concatemers with different promoter-enhancer interacting patterns for the two genes. \u003cstrong\u003eG\u003c/strong\u003e, hypothesized haplotype-specific high-order interaction models for genes subjected to (left) and escaped (right) from XCI.\u003c/p\u003e","description":"","filename":"image7.tiff.png","url":"https://assets-eu.researchsquare.com/files/rs-3486881/v1/d7a5d9e1191d2632fe3a7f88.png"},{"id":77753274,"identity":"da4eac16-9372-4b31-804a-af8e1602a951","added_by":"auto","created_at":"2025-03-05 08:06:01","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3871681,"visible":true,"origin":"","legend":"","description":"","filename":"20241024Dip3Dmanuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3486881/v1_covered_5a4af0e4-bcab-44ca-b08d-8dae86b8ab3f.pdf"},{"id":67806829,"identity":"d62e4a2d-4e27-49b3-8ce4-1ec5b1455154","added_by":"auto","created_at":"2024-10-30 01:40:31","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":5927896,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Tables\u003c/p\u003e","description":"","filename":"20241024SupplementaryTables.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-3486881/v1/13116a7d4924f29110e65ec2.xlsx"},{"id":67807064,"identity":"bb465f3c-8e90-41c6-af6c-5256db517775","added_by":"auto","created_at":"2024-10-30 01:48:31","extension":"pdf","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":1364994,"visible":true,"origin":"","legend":"","description":"","filename":"20241024Dip3DSupplementaryNotes.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3486881/v1/fa323cb5213c1f15c82193b9.pdf"},{"id":67806833,"identity":"c1bd4e26-c15d-409e-b1a2-26170762cf50","added_by":"auto","created_at":"2024-10-30 01:40:31","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":7998471,"visible":true,"origin":"","legend":"","description":"","filename":"20241024SupplementaryFigures.docx","url":"https://assets-eu.researchsquare.com/files/rs-3486881/v1/e793b83a36358e76d03a5349.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Reconstruction of Diploid High-Order 3D Genome Interactions from Long Noisy Concatemers","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-3486881/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3486881/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Differential high-order chromatin interactions between homologous chromosomes play pivotal roles in many biological processes. However, their elucidation has been hindered by technical difficulties. Traditional 3C methods mainly expose 2-way interactions and offer limited haplotype information. In response, we addressed challenges in harnessing merely nanopore high-order concatemer (Pore-C) reads to delineate diploid high-order chromatin interactions. By training a cutting-edge deep learning model and making statistical analysis, we achieved superior SNV calling and haplo-tagging for noisy short monomers. Learning the haplotype characteristics of high-order concatemers allowed us to devise a progressive haplotype imputation strategy, which elevated the haplotype informative Pore-C contact rate by 14.1-fold to 76% in the HG001 cell line, eclipsing Hi-C's rate by over ten times. Overall, our diploid 3D genome interactions surpassed conventional methods in noise reduction and contact distribution uniformity, with over one magnitude advantages in haplotype informative contact density and genomic coverage rate. Dip3D enabled the unveiling of haplotype high-order interactions in many previously overlooked genomic regions and the investigation of their relationship with allele-specific expression, especially concerning X-chromosome inactivation. Our approach stands out as a robust and accessible solution for the high-quality reconstruction of diploid high-order 3D genome interactions.","manuscriptTitle":"Reconstruction of Diploid High-Order 3D Genome Interactions from Long Noisy Concatemers","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-10-30 01:40:26","doi":"10.21203/rs.3.rs-3486881/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-structural-and-molecular-biology","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"nsmb","sideBox":"Learn more about [Nature Structural \u0026 Molecular Biology](http://www.nature.com/nsmb/)","snPcode":"","submissionUrl":"","title":"Nature Structural \u0026 Molecular Biology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Research","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"ad40d91d-b699-462e-8a11-7365c779131c","owner":[],"postedDate":"October 30th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":39506457,"name":"Biological sciences/Computational biology and bioinformatics/High-throughput screening"},{"id":39506458,"name":"Biological sciences/Genetics/Genomics/Epigenomics"}],"tags":[],"updatedAt":"2025-03-05T08:05:39+00:00","versionOfRecord":{"articleIdentity":"rs-3486881","link":"https://doi.org/10.1038/s41594-025-01512-w","journal":{"identity":"nature-structural-and-molecular-biology","isVorOnly":false,"title":"Nature Structural \u0026 Molecular Biology"},"publishedOn":"2025-03-04 05:00:00","publishedOnDateReadable":"March 4th, 2025"},"versionCreatedAt":"2024-10-30 01:40:26","video":"","vorDoi":"10.1038/s41594-025-01512-w","vorDoiUrl":"https://doi.org/10.1038/s41594-025-01512-w","workflowStages":[]},"version":"v1","identity":"rs-3486881","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3486881","identity":"rs-3486881","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}