EpiMCI: Predicting Multi-Way Chromatin Interactions from Epigenomic Signals

The recently emerging high-throughput Pore-C (HiPore-C) can identify whole-genome high-order chromatin multi-way interactions with an ultra-high output, contributing to deciphering three-dimensional (3D) genome organization. However, it also brings new challenges to relevant data analysis. To alleviate this problem, we proposed the EpiMCI, a model for multi-way chromatin interaction prediction based on a hypergraph neural network with epigenomic signals as the input. The EpiMCI integrated separate hyperedge representations with coupling hyperedge information and obtained AUCs of 0.981 and 0.984 in the GM12878 and K562 datasets, respectively, which outperformed the current available method. Moreover, the EpiMCI can be applied to denoise the HiPore-C data and improve the data quality efficiently. Furthermore, the vertex embeddings extracted from the EpiMCI reflected the global chromatin architecture accurately. The principal component analysis suggested that it was well aligned with the activities of genomic regions at the chromatin compartment level. Taken together, the EpiMCI can accurately predict multi-way chromatin interactions and can be applied to studies relying on chromatin architecture.

[1]  Chunhui Hou,et al.  High-throughput Pore-C reveals the single-allele topology and cell type-specificity of 3D genome folding , 2023, Nature Communications.

[2]  Yue Gao,et al.  HGNN+: General Hypergraph Neural Networks , 2022, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[3]  Julie M. Behr,et al.  Identifying synergistic high-order 3D chromatin conformations from genome-scale nanopore concatemer sequencing , 2022, Nature Biotechnology.

[4]  William Stafford Noble,et al.  Epiphany: predicting Hi-C contact maps from 1D epigenomic signals , 2021, bioRxiv.

[5]  I. Rajapakse,et al.  Deciphering multi-way interactions in the human genome , 2021, bioRxiv.

[6]  Tianqi Quan,et al.  A model-based collaborate filtering algorithm based on stacked AutoEncoder , 2021, Neural Computing and Applications.

[7]  Ying Wang,et al.  TAERT: Triple-Attentional Explainable Recommendation with Temporal Convolutional Network , 2021, Inf. Sci..

[8]  D. Higgs,et al.  The relationship between genome structure and function , 2020, Nature Reviews Genetics.

[9]  T. Misteli.  The Self-Organizing Genome: Principles of Genome Architecture and Function , 2020, Cell.

[10]  Mikhail G. Dozmorov,et al.  preciseTAD: a transfer learning framework for 3D domain boundary prediction at base-pair resolution , 2020, bioRxiv.

[11]  Sheng Li,et al.  Graph embedding and unsupervised learning predict genomic sub-compartments from HiC chromatin interaction data , 2020, Nature Communications.

[12]  Jian Ma,et al.  Probing multi-way chromatin interaction with hypergraph representation learning , 2020, bioRxiv.

[13]  A. Pombo,et al.  Methods for mapping 3D chromosome architecture , 2019, Nature Reviews Genetics.

[14]  Kyle Xiong,et al.  Revealing Hi-C subcompartments by imputing inter-chromosomal chromatin interactions , 2019, Nature Communications.

[15]  Wei Xie,et al.  The role of 3D genome organization in development and cell differentiation , 2019, Nature Reviews Molecular Cell Biology.

[16]  Bing Ren,et al.  A Compendium of Promoter-Centered Long-Range Chromatin Interactions in the Human Genome , 2019, Nature Genetics.

[17]  Guillaume J. Filion,et al.  Transcription factors and 3D genome conformation in cell-fate decisions , 2019, Nature.

[18]  Chia-Lin Wei,et al.  Multiplex chromatin interactions with single-molecule precision , 2019, Nature.

[19]  Simona Bianco,et al.  Single-allele chromatin interactions identify regulatory hubs in dynamic compartmentalized domains , 2018, Nature Genetics.

[20]  Britta A. M. Bouwman,et al.  Enhancer hubs and loop collisions identified from single-allele topologies , 2018, Nature Genetics.

[21]  B. Tabak,et al.  Higher-Order Inter-chromosomal Hubs Shape 3D Genome Organization in the Nucleus , 2018, Cell.

[22]  Siheng Zhang,et al.  Digestion-ligation-only Hi-C is an efficient and cost-effective method for chromosome conformation capture , 2018, Nature Genetics.

[23]  Erez Lieberman Aiden,et al.  Cohesin Loss Eliminates All Loop Domains , 2017, Cell.

[24]  Kevin Y. Yip,et al.  Reconstruction of enhancer–target networks in 935 samples of human primary cells, tissues and cell lines , 2017, Nature Genetics.

[25]  William Stafford Noble,et al.  GenomeDISCO: A concordance score for chromosome conformation capture experiments using random walks on contact map graphs , 2017, bioRxiv.

[26]  Erez Lieberman Aiden,et al.  De novo prediction of human chromosome structures: Epigenetic marking patterns encode genome architecture , 2017, Proceedings of the National Academy of Sciences.

[27]  Lukasz Kaiser,et al.  Attention is All you Need , 2017, NIPS.

[28]  Niels Galjart,et al.  Cohesin is positioned in mammalian genomes by transcription, CTCF and Wapl , 2017, Nature.

[29]  S. Q. Xie,et al.  Complex multi-enhancer contacts captured by Genome Architecture Mapping (GAM) , 2017, Nature.

[30]  Barnabás Póczos,et al.  Predicting enhancer-promoter interaction from genomic sequence with deep neural networks , 2016, bioRxiv.

[31]  Jure Leskovec,et al.  Higher-order organization of complex networks , 2016, Science.

[32]  Neva C. Durand,et al.  Juicer Provides a One-Click System for Analyzing Loop-Resolution Hi-C Experiments. , 2016, Cell systems.

[33]  Xavier Bresson,et al.  Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering , 2016, NIPS.

[34]  K. Pollard,et al.  Enhancer–promoter interactions are encoded by complex genomic signatures on looping chromatin , 2016, Nature Genetics.

[35]  Wei Wang,et al.  Constructing 3D interaction maps from 1D epigenomes , 2016, Nature Communications.

[36]  Alireza F. Siahpirani,et al.  A predictive modeling approach for cell line-specific long-range regulatory interactions , 2015, Nucleic acids research.

[37]  P. Fraser,et al.  Comparison of Hi-C results using in-solution versus in-nucleus ligation , 2015, Genome Biology.

[38]  Michael Q. Zhang,et al.  CRISPR Inversion of CTCF Sites Alters Genome Topology and Enhancer/Promoter Function , 2015, Cell.

[39]  A. Visel,et al.  Disruptions of Topological Chromatin Domains Cause Pathogenic Rewiring of Gene-Enhancer Interactions , 2015, Cell.

[40]  J. Dekker,et al.  Condensin-Driven Remodeling of X-Chromosome Topology during Dosage Compensation , 2015, Nature.

[41]  Wei Wang,et al.  Computational schemes for the prediction and annotation of enhancers from epigenomic assays. , 2015, Methods.

[42]  Neva C. Durand,et al.  A 3D Map of the Human Genome at Kilobase Resolution Reveals Principles of Chromatin Looping , 2014, Cell.

[43]  K. Tan,et al.  Global view of enhancer–promoter interactome in human cells , 2014, Proceedings of the National Academy of Sciences.

[44]  Simon J. Doran,et al.  Stacked Autoencoders for Unsupervised Feature Learning and Multiple Organ Detection in a Pilot Study Using 4D Patient Data , 2013, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[45]  Zhaohui S. Qin,et al.  Gene density, transcription, and insulators contribute to the partition of the Drosophila genome into physical domains. , 2012, Molecular cell.

[46]  Jesse R. Dixon,et al.  Topological Domains in Mammalian Genomes Identified by Analysis of Chromatin Interactions , 2012, Nature.

[47]  Y. Ruan,et al.  ChIP‐based methods for the identification of long‐range chromatin interactions , 2009, Journal of cellular biochemistry.

[48]  Yoshua Bengio,et al.  Greedy Layer-Wise Training of Deep Networks , 2006, NIPS.

[49]  Gao Daqi,et al.  Classification methodologies of multilayer perceptrons with sigmoid activation functions , 2005, Pattern Recognit..

[50]  I. Amit,et al.  Comprehensive mapping of long-range interactions reveals folding principles of the human genome. , 2009, Science.

[51]  Shie Mannor,et al.  A Tutorial on the Cross-Entropy Method , 2005, Ann. Oper. Res..