Machine and Deep Learning Methods for Predicting 3D Genome Organization.

Three-dimensional (3D) chromatin interactions, such as enhancer-promoter interactions (EPIs), loops, topologically associating domains (TADs), and A/B compartments, play critical roles in a wide range of cellular processes by regulating gene expression. Recent development of chromatin conformation capture technologies has enabled genome-wide profiling of various 3D structures, even with single cells. However, current catalogs of 3D structures remain incomplete and unreliable due to differences in technology, tools, and low data resolution. Machine learning methods have emerged as an alternative to obtain missing 3D interactions and/or improve resolution. Such methods frequently use genome annotation data (ChIP-seq, DNAse-seq, etc.), DNA sequencing information (k-mers and transcription factor binding site (TFBS) motifs), and other genomic properties to learn the associations between genomic features and chromatin interactions. In this review, we discuss computational tools for predicting three types of 3D interactions (EPIs, chromatin interactions, and TAD boundaries) and analyze their pros and cons. We also point out obstacles to the computational prediction of 3D interactions and suggest future research directions.

Read Mapping for Hi-C Analysis.

Hi-C is a popular ligation-based technique to detect 3D physical chromosome structure within the nucleus using cross-linking and next-generation sequencing. As an unbiased genome-wide assay based on chromosome conformation capture, it provides rich insights into chromosome structure, dynamic chromosome folding and interactions, and the regulatory state of a cell. Bioinformatics analyses of Hi-C data require dedicated protocols as most genome alignment tools assume that both paired-end reads will map to the same chromosome, resulting in large two-dimensional matrices as processed data. Here, we outline the necessary steps to generate high-quality aligned Hi-C data by separately mapping each read while correcting for biases from restriction enzyme digests. We introduce our own custom open-source pipeline, which enables users to select an aligner of their choosing with high accuracy and performance. This enables users to generate high-resolution datasets with fast turnaround and fewer unmapped reads. Finally, we discuss recent innovations in experimental techniques, bioinformatics techniques, and their applications in clinical testing for diagnostics.

Exploring Contact Distance Distributions with Google Colaboratory.

Hi-C and Micro-C are the three-dimensional (3D) genome assays that use high-throughput sequencing. In the analysis, the sequenced paired-end reads are mapped to a reference genome to generate a two-dimensional contact matrix for identifying topologically associating domains (TADs), chromatin loops, and chromosomal compartments. On the other hand, the distance distribution of the paired-end mapped reads also provides insight into the 3D genome structure by highlighting global contact frequency patterns at distances indicative of loops, TADs, and compartments. This chapter presents a basic workflow for visualizing and analyzing contact distance distributions from Hi-C data. The workflow can be run on Google Colaboratory, which provides a ready-to-use Python environment accessible through a web browser. The notebook that demonstrates the workflow is available in the GitHub repository at https://github.com/rnakato/Springer_contact_distance_plot.

Systematic Inference of Multi-scale Chromatin Sub-compartments Using Calder2.

The compartmentalization of chromatin reflects its underlying biological activities. Inferring chromatin sub-compartments using Hi-C data is challenged by data resolution constraints. Consequently, comprehensive characterizations of sub-compartments have been limited to a select number of Hi-C experiments, with systematic comparisons across a wide range of tissues and conditions still lacking. Our original Calder algorithm marked a significant advancement in this field, enabling the identification of multi-scale sub-compartments at various data resolutions and facilitating the inference and comparison of chromatin architecture in over 100 datasets. Building on this foundation, we introduce Calder2, an updated version of Calder that brings notable improvements. These include expanded support for a wider array of genomes and organisms, an optimized bin size selection approach for more accurate chromatin compartment detection, and extended support for input and output formats. Calder2 thus stands as a refined analysis tool, significantly advancing genome-wide studies of 3D chromatin architecture and its functional implications.

CWL-Based Analysis Pipeline for Hi-C Data: From FASTQ Files to Matrices.

Over a decade has passed since the development of the Hi-C method for genome-wide analysis of 3D genome organization. Hi-C utilizes next-generation sequencing (NGS) technology to generate large-scale chromatin interaction data, which has accumulated across a diverse range of species and cell types, particularly in eukaryotes. There is thus a growing need to streamline the process of Hi-C data analysis to utilize these data sets effectively. Hi-C generates data that are much larger compared to other NGS techniques such as chromatin immunoprecipitation sequencing (ChIP-seq) or RNA-seq, making the data reanalysis process computationally expensive. In an effort to bridge this resource gap, the 4D Nucleome (4DN) Data Portal has reanalyzed approximately 600 Hi-C data sets, allowing users to access and utilize the analyzed data. In this chapter, we provide detailed instructions for the implementation of the common workflow language (CWL)-based Hi-C analysis pipeline adopted by the 4DN Data Portal ecosystem. This reproducible and portable pipeline generates standard Hi-C contact matrices in formats such as .hic or .mcool from FASTQ files. It enables users to output their own Hi-C data in the same format as those registered in the 4DN Data portal, facilitating comparative analysis using data registered in the portal. Our custom-made scripts are available on GitHub at https://github.com/kuzobuta/4dn_cwl_pipeline .

Step-by-Step Protocol to Generate Hi-C Contact Maps Using the rfy_hic2 Pipeline.

The Hi-C method has emerged as an indispensable tool for analyzing the 3D organization of the genome, becoming increasingly accessible and frequently utilized in chromatin research. To effectively leverage 3D genomics data obtained through advanced technologies, it is crucial to understand what processes are undertaken and what aspects require special attention within the bioinformatics pipeline. This protocol aims to demystify the Hi-C data analysis process for field newcomers. In a step-by-step manner, we describe how to process Hi-C data, from the initial sequencing of the Hi-C library to the final visualization of Hi-C contact data as heatmaps. Each step of the analysis is clearly explained to ensure an understanding of the procedures and their objectives. By the end of this chapter, readers will be equipped with the knowledge to transform raw Hi-C reads into informative visual representations, facilitating a deeper comprehension of the spatial genomic structures critical to cellular functions.

Supervised Chromatin Loop Detection Using Peakachu Version 2.

Peakachu is a supervised-learning-based approach that identifies chromatin loops from chromatin contact data. Here, we present Peakachu version 2, an updated version that significantly improves extensibility, usability, and computational efficiency compared to its predecessor. It features pretrained models tailored for a wide range of experimental platforms, such as Hi-C, Micro-C, ChIA-PET, HiChIP, HiCAR, and TrAC-loop. This chapter offers a step-by-step tutorial guiding users through the process of training Peakachu models from scratch and utilizing pretrained models to predict chromatin loops across various platforms.

Integrative Modeling of 3D Genome Organization by Bayesian Molecular Dynamics Simulations with Hi-C Metainference.

Polymer modeling has been playing an increasingly important role in complementing 3D genome experiments, both to aid their interpretation and to reveal the underlying molecular mechanisms. This chapter illustrates an application of Hi-C metainference, a Bayesian approach to explore the 3D organization of a target genomic region by integrating experimental contact frequencies into a prior model of chromatin. The method reconstructs the conformational ensemble of the target locus by combining molecular dynamics simulation and Monte Carlo sampling from the posterior probability distribution given the data. Using prior chromatin models at both 1 kb and nucleosome resolution, we apply this approach to a 30 kb locus of mouse embryonic stem cells consisting of two well-defined domains linking several gene promoters together. Retaining the advantages of both physics-based and data-driven strategies, Hi-C metainference can provide an experimentally consistent representation of the system while at the same time retaining molecular details necessary to derive physical insights.

Prediction of Enhancer-Gene Interactions Using Chromatin-Conformation Capture and Epigenome Data Using STARE.

Disentangling the relationship of enhancers and genes is an ongoing challenge in epigenomics. We present STARE, our software to quantify the strength of enhancer-gene interactions based on enhancer activity and chromatin contact data. It implements the generalized Activity-by-Contact (gABC) score, which allows predicting putative target genes of candidate enhancers over any desired genomic distance. The only requirement for its application is a measurement of enhancer activity. In addition to regulatory interactions, STARE calculates transcription factor (TF) affinities on gene level. We illustrate its usage on a public single-cell data set of the human heart by predicting regulatory interactions on cell type level, by giving examples on how to integrate them with other data modalities, and by constructing TF affinity matrices.

4D Genome Analysis Using PHi-C2.

Hi-C methods reveal 3D genome features but lack correspondence to dynamic chromatin behavior. PHi-C2, Python software, addresses this gap by transforming Hi-C data into polymer models. After the optimization algorithm, it enables us to calculate 3D conformations and conduct dynamic simulations, providing insights into chromatin dynamics, including the mean-squared displacement and rheological properties. This chapter introduces PHi-C2 usage, offering a tutorial for comprehensive 4D genome analysis.

Three-Dimensional Simulation of Whole-Genome Structuring Through the Transition from Anaphase to Interphase.

In order to analyze the three-dimensional genome architecture, it is important to simulate how the genome is structured through the cell cycle progression. In this chapter, we present the usage of our computation codes for simulating how the human genome is formed as the cell transforms from anaphase to interphase. We do not use the global Hi-C data as an input into the genome simulation but represent all chromosomes as linear polymers annotated by the neighboring region contact index (NCI), which classifies the A/B type of each local chromatin region. The simulated mitotic chromosomes heterogeneously expand upon entry to the G1 phase, which induces phase separation of A and B chromatin regions, establishing chromosome territories, compartments, and lamina and nucleolus associations in the interphase nucleus. When the appropriate one-dimensional chromosomal annotation is possible, using the protocol of this chapter, one can quantitatively simulate the three-dimensional genome structure and dynamics of human cells of interest.

Exploring Cohesin-Related Multiomics Information in the Web Browser.

Cohesin is a protein complex that plays a key role in regulating chromosome structure and gene expression. While next-generation sequencing technologies have provided extensive information on various aspects of cohesin, integrating and exploring the vast datasets associated with cohesin are not straightforward. CohesinDB ( https://cohesindb.iqb.u-tokyo.ac.jp ) offers a web-based interface for browsing, searching, analyzing, visualizing, and downloading comprehensive multiomics cohesin information in human cells. In this protocol, we introduce how to utilize CohesinDB to facilitate research on transcriptional regulation and chromatin organization.

Methods for Genome-Wide Chromatin Interaction Analysis.

Recent analyses revealed the essential function of chromatin structure in maintaining and regulating genomic information. Advancements in microscopy, nuclear structure observation techniques, and the development of methods utilizing next-generation sequencers (NGSs) have significantly progressed these discoveries. Methods utilizing NGS enable genome-wide analysis, which is challenging with microscopy, and have elucidated concepts of important chromatin structures such as a loop structure, a domain structure called topologically associating domains (TADs), and compartments. In this chapter, I introduce chromatin interaction techniques using NGS and outline the principles and features of each method.

Investigating the cis-regulatory basis of C3 and C4 photosynthesis in grasses at single-cell resolution.

While considerable knowledge exists about the enzymes pivotal for C4 photosynthesis, much less is known about the cis-regulation important for specifying their expression in distinct cell types. Here, we use single-cell-indexed ATAC-seq to identify cell-type-specific accessible chromatin regions (ACRs) associated with C4 enzymes for five different grass species. This study spans four C4 species, covering three distinct photosynthetic subtypes: Zea mays and Sorghum bicolor (NADP-dependent malic enzyme), Panicum miliaceum (NAD-dependent malic enzyme), Urochloa fusca (phosphoenolpyruvate carboxykinase), along with the C3 outgroup Oryza sativa. We studied the cis-regulatory landscape of enzymes essential across all C4 species and those unique to C4 subtypes, measuring cell-type-specific biases for C4 enzymes using chromatin accessibility data. Integrating these data with phylogenetics revealed diverse co-option of gene family members between species, showcasing the various paths of C4 evolution. Besides promoter proximal ACRs, we found that, on average, C4 genes have two to three distal cell-type-specific ACRs, highlighting the complexity and divergent nature of C4 evolution. Examining the evolutionary history of these cell-type-specific ACRs revealed a spectrum of conserved and novel ACRs, even among closely related species, indicating ongoing evolution of cis-regulation at these C4 loci. This study illuminates the dynamic and complex nature of cis-regulatory elements evolution in C4 photosynthesis, particularly highlighting the intricate cis-regulatory evolution of key loci. Our findings offer a valuable resource for future investigations, potentially aiding in the optimization of C3 crop performance under changing climatic conditions.

Timely lagging strand maturation relies on Ubp10 deubiquitylase-mediated PCNA dissociation from replicating chromatin.

Synthesis and maturation of Okazaki Fragments is an incessant and highly efficient metabolic process completing the synthesis of the lagging strands at replication forks during S phase. Accurate Okazaki fragment maturation (OFM) is crucial to maintain genome integrity and, therefore, cell survival in all living organisms. In eukaryotes, OFM involves the consecutive action of DNA polymerase Pol ∂, 5' Flap endonuclease Fen1 and DNA ligase I, and constitutes the best example of a sequential process coordinated by the sliding clamp PCNA. For OFM to occur efficiently, cooperation of these enzymes with PCNA must be highly regulated. Here, we present evidence of a role for the K164-PCNA-deubiquitylase Ubp10 in the maturation of Okazaki fragments in the budding yeast Saccharomyces cerevisiae. We show that Ubp10 associates with lagging-strand DNA synthesis machineries on replicating chromatin to ensure timely ligation of Okazaki fragments by promoting PCNA dissociation from chromatin requiring lysine 164 deubiquitylation.

PHYTOCHROME-INTERACTING FACTOR 7 and RELATIVE OF EARLY FLOWERING 6 act in shade avoidance memory in Arabidopsis.

Shade avoidance helps plants maximize their access to light for growth under crowding. It is unknown, however, whether a priming shade avoidance mechanism exists that allows plants to respond more effectively to successive shade conditions. Here, we show that the shade-intolerant plant Arabidopsis can remember a first experienced shade event and respond more efficiently to the next event on hypocotyl elongation. The transcriptional regulator PHYTOCHROME-INTERACTING FACTOR 7 (PIF7) and the histone H3K27-demethylase RELATIVE OF EARLY FLOWERING 6 (REF6) are identified as being required for this shade avoidance memory. RNA-sequencing analysis reveals that shade induction of shade-memory-related genes is impaired in the pif7 and ref6 mutants. Based on the analyses of enrichments of H3K27me3, REF6 and PIF7, we find that priming shade treatment induces PIF7 accumulation, which further recruits REF6 to demethylate H3K27me3 on the chromatin of certain shade-memory-related genes, leading to a state poised for their transcription. Upon a second shade treatment, enhanced shade-mediated inductions of these genes result in stronger hypocotyl growth responses. We conclude that the transcriptional memory mediated by epigenetic modification plays a key role in the ability of primed plants to remember previously experienced shade and acquire enhanced responses to recurring shade conditions.

USP50 suppresses alternative RecQ helicase use and deleterious DNA2 activity during replication.

Mammalian DNA replication relies on various DNA helicase and nuclease activities to ensure accurate genetic duplication, but how different helicase and nuclease activities are properly directed remains unclear. Here, we identify the ubiquitin-specific protease, USP50, as a chromatin-associated protein required to promote ongoing replication, fork restart, telomere maintenance, cellular survival following hydroxyurea or pyridostatin treatment, and suppression of DNA breaks near GC-rich sequences. We find that USP50 supports proper WRN-FEN1 localisation at or near stalled replication forks. Nascent DNA in cells lacking USP50 shows increased association of the DNA2 nuclease and RECQL4 and RECQL5 helicases and replication defects in cells lacking USP50, or FEN1 are driven by these proteins. Consequently, suppression of DNA2 or RECQL4/5 improves USP50-depleted cell resistance to agents inducing replicative stress and restores telomere stability. These data define an unexpected regulatory protein that promotes the balance of helicase and nuclease use at ongoing and stalled replication forks.

The emerging role of SARS-CoV-2 nonstructural protein 1 (nsp1) in epigenetic regulation of host gene expression.

Infection with the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes widespread changes in epigenetic modifications and chromatin architecture in the host cell. Recent evidence suggests that SARS-CoV-2 nonstructural protein 1 (nsp1) plays an important role in driving these changes. Previously thought to be primarily involved in host translation shutoff and cellular mRNA degradation, nsp1 has now been shown to be a truly multifunctional protein that affects host gene expression at multiple levels. The functions of nsp1 are surprisingly diverse and include not only the downregulation of cellular mRNA translation and stability, but also the inhibition of mRNA export from the nucleus, the suppression of host immune signaling, and, most recently, the epigenetic regulation of host gene expression. In this review, we first summarize the current knowledge on SARS-CoV-2-induced changes in epigenetic modifications and chromatin structure. We then focus on the role of nsp1 in epigenetic reprogramming, with a particular emphasis on the silencing of immune-related genes. Finally, we discuss potential molecular mechanisms underlying the epigenetic functions of nsp1 based on evidence from SARS-CoV-2 interactome studies.

Clusters of lineage-specific genes are anchored by ZNF274 in repressive perinucleolar compartments.

Long known as the site of ribosome biogenesis, the nucleolus is increasingly recognized for its role in shaping three-dimensional (3D) genome organization. Still, the mechanisms governing the targeting of selected regions of the genome to nucleolus-associated domains (NADs) remain enigmatic. Here, we reveal the essential role of ZNF274, a SCAN-bearing member of the Krüppel-associated box (KRAB)-containing zinc finger protein (KZFP) family, in sequestering lineage-specific gene clusters within NADs. Ablation of ZNF274 triggers transcriptional activation across entire genomic neighborhoods-encompassing, among others, protocadherin and KZFP-encoding genes-with loss of repressive chromatin marks, altered the 3D genome architecture and de novo CTCF binding. Mechanistically, ZNF274 anchors target DNA sequences at the nucleolus and facilitates their compartmentalization via a previously uncharted function of the SCAN domain. Our findings illuminate the mechanisms underlying NAD organization and suggest that perinucleolar entrapment into repressive hubs constrains the activation of tandemly arrayed genes to enable selective expression and modulate cell differentiation programs during development.