7C: Computational Chromosome Conformation Capture by Correlation of ChIP-seq at CTCF motifs

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.

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, 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, TAD boundaries) and analyze their pros and cons. We also point out obstacles of computational prediction of 3D interactions and suggest future research directions.

PAXIP1 and STAG2 converge to maintain 3D genome architecture and facilitate promoter/enhancer contacts to enable stress hormone-dependent transcription

How steroid hormone receptors (SHRs) regulate transcriptional activity remains partly understood. Upon activation, SHRs bind the genome together with a co-regulator repertoire, crucial to induce gene expression. However, it remains unknown which components of the SHR-recruited co-regulator complex are essential to drive transcription following hormonal stimuli. Through a FACS-based genome-wide CRISPR screen, we functionally dissected the Glucocorticoid Receptor (GR) complex. We describe a functional cross-talk between PAXIP1 and the cohesin subunit STAG2, critical for regulation of gene expression by GR. Without altering the GR cistrome, PAXIP1 and STAG2 depletion alter the GR transcriptome, by impairing the recruitment of 3D-genome organization proteins to the GR complex. Importantly, we demonstrate that PAXIP1 is required for stability of cohesin on chromatin, its localization to GR-occupied sites, and maintenance of enhancer-promoter interactions. In lung cancer, where GR acts as tumor suppressor, PAXIP1/STAG2 loss enhances GR-mediated tumor suppressor activity by modifying local chromatin interactions. All together, we introduce PAXIP1 and STAG2 as novel co-regulators of GR, required to maintain 3D-genome architecture and drive the GR transcriptional programme following hormonal stimuli.

Inferring CTCF-binding patterns and anchored loops across human tissues and cell types

CCCTC-binding factor (CTCF) is a transcription regulator with a complex role in gene regulation. The recognition and effects of CTCF on DNA sequences, chromosome barriers, and enhancer blocking are not well understood. Existing computational tools struggle to assess the regulatory potential of CTCF-binding sites and their impact on chromatin loop formation. Here we have developed a deep-learning model, DeepAnchor, to accurately characterize CTCF binding using high-resolution genomic/epigenomic features. This has revealed distinct chromatin and sequence patterns for CTCF-mediated insulation and looping. An optimized implementation of a previous loop model based on DeepAnchor score excels in predicting CTCF-anchored loops. We have established a compendium of CTCF-anchored loops across 52 human tissue/cell types, and this suggests that genomic disruption of these loops could be a general mechanism of disease pathogenesis. These computational models and resources can help investigate how CTCF-mediated cis-regulatory elements shape context-specific gene regulation in cell development and disease progression.

Krüppel-Like Factors Orchestrate Endothelial Gene Expression Through Redundant and Non-Redundant Enhancer Networks

Background Proper function of endothelial cells is critical for vascular integrity and organismal survival. Studies over the past 2 decades have identified 2 members of the KLF (Krüppel-like factor) family of proteins, KLF2 and KLF4, as nodal regulators of endothelial function. Strikingly, inducible postnatal deletion of both KLF2 and KLF4 resulted in widespread vascular leak, coagulopathy, and rapid death. Importantly, while transcriptomic studies revealed profound alterations in gene expression, the molecular mechanisms underlying these changes remain poorly understood. Here, we seek to determine mechanisms of KLF2 and KLF4 transcriptional control in multiple vascular beds to further understand their roles as critical endothelial regulators. Methods and Results We integrate chromatin occupancy and transcription studies from multiple transgenic mouse models to demonstrate that KLF2 and KLF4 have overlapping yet distinct binding patterns and transcriptional targets in heart and lung endothelium. Mechanistically, KLFs use open chromatin regions in promoters and enhancers and bind in context-specific patterns that govern transcription in microvasculature. Importantly, this occurs during homeostasis in vivo without additional exogenous stimuli. Conclusions Together, this work provides mechanistic insight behind the well-described transcriptional and functional heterogeneity seen in vascular populations, while also establishing tools into exploring microvascular endothelial dynamics in vivo.

TF-COMB - discovering grammar of transcription factor binding sites

Cooperativity between transcription factors is important to regulate target gene expression. In particular, the binding grammar of TFs in relation to each other, as well as in the context of other genomic elements, is crucial for TF functionality. However, tools to easily uncover co-occurrence between DNA-binding proteins, and investigate the regulatory modules of TFs, are limited. Here we present TF-COMB (Transcription Factor Co-Occurrence using Market Basket analysis) - a tool to investigate co-occurring TFs and binding grammar within regulatory regions. We found that TF-COMB can accurately identify known co-occurring TFs from ChIP-seq data, as well as uncover preferential localization to other genomic elements. With the use of ATAC-seq footprinting and TF motif locations, we found that TFs exhibit both preferred orientation and distance in relation to each other, and that these are biologically significant. Finally, we extended the analysis to not only investigate individual TF pairs, but also TF pairs in the context of networks, which enabled the investigation of TF complexes and TF hubs. In conclusion, TF-COMB is a flexible tool to investigate various aspects of TF binding grammar.

A machine learning framework for the prediction of chromatin folding in Drosophila using epigenetic features

Technological advances have lead to the creation of large epigenetic datasets, including information about DNA binding proteins and DNA spatial structure. Hi-C experiments have revealed that chromosomes are subdivided into sets of self-interacting domains called Topologically Associating Domains (TADs). TADs are involved in the regulation of gene expression activity, but the mechanisms of their formation are not yet fully understood. Here, we focus on machine learning methods to characterize DNA folding patterns in Drosophila based on chromatin marks across three cell lines. We present linear regression models with four types of regularization, gradient boosting, and recurrent neural networks (RNN) as tools to study chromatin folding characteristics associated with TADs given epigenetic chromatin immunoprecipitation data. The bidirectional long short-term memory RNN architecture produced the best prediction scores and identified biologically relevant features. Distribution of protein Chriz (Chromator) and histone modification H3K4me3 were selected as the most informative features for the prediction of TADs characteristics. This approach may be adapted to any similar biological dataset of chromatin features across various cell lines and species. The code for the implemented pipeline, Hi-ChiP-ML, is publicly available: https://github.com/MichalRozenwald/Hi-ChIP-ML.

Quantitative prediction of enhancer-promoter interactions

Recent experimental and computational efforts have provided large data sets describing three-dimensional organization of mouse and human genomes and showed the interconnection between the expression profile, epigenetic state, and spatial interactions of loci. These interconnections were utilized to infer the spatial organization of chromatin, including enhancer–promoter contacts, from one-dimensional epigenetic marks. Here, we show that the predictive power of some of these algorithms is overestimated due to peculiar properties of the biological data. We propose an alternative approach, which provides high-quality predictions of chromatin interactions using information on gene expression and CTCF-binding alone. Using multiple metrics, we confirmed that our algorithm could efficiently predict the three-dimensional architecture of both normal and rearranged genomes.