Search

Gene regulation is inherently multiscale, but scale-adaptive machine learning methods that fully exploit this property in single-nucleus accessibility data are still lacking. Here, we develop ChromatinHD, a pair of scale-adaptive models that uses the raw accessibility data, without peak-calling or windows, to link regions to gene expression and determine differentially accessible chromatin. We show how ChromatinHD consistently outperforms existing peak and window-based approaches and find that this is due to a large number of uniquely captured, functional accessibility changes within and outside of putative cis-regulatory regions. Furthermore, ChromatinHD can delineate collaborating regulatory regions, including their preferential genomic conformations, that drive gene expression. Finally, our models also use changes in ATAC-seq fragment lengths to identify dense binding of transcription factors, a feature not captured by footprinting methods. Altogether, ChromatinHD, available at https://chromatinhd.org , is a suite of computational tools that enables a data-driven understanding of chromatin accessibility at various scales and how it relates to gene expression., Competing Interests: Competing interests: The authors declare no competing interests., (© 2024. The Author(s).)

Identifying cell-type-specific 3D chromatin interactions between regulatory elements can help decipher gene regulation and interpret disease-associated non-coding variants. However, achieving this resolution with current 3D genomics technologies is often infeasible given limited input cell numbers. We therefore present ChromaFold, a deep learning model that predicts 3D contact maps, including regulatory interactions, from single-cell ATAC sequencing (scATAC-seq) data alone. ChromaFold uses pseudobulk chromatin accessibility, co-accessibility across metacells, and a CTCF motif track as inputs and employs a lightweight architecture to train on standard GPUs. Trained on paired scATAC-seq and Hi-C data in human samples, ChromaFold accurately predicts the 3D contact map and peak-level interactions across diverse human and mouse test cell types. Compared to leading contact map prediction models that use ATAC-seq and CTCF ChIP-seq, ChromaFold achieves state-of-the-art performance using only scATAC-seq. Finally, fine-tuning ChromaFold on paired scATAC-seq and Hi-C in a complex tissue enables deconvolution of chromatin interactions across cell subpopulations., (© 2024. The Author(s).)

Histone modifications can regulate transcription epigenetically by marking specific genomic loci, which can be mapped using chromatin immunoprecipitation sequencing (ChIP-seq). Here we present QHistone, a predictive database of 1534 ChIP-seqs from 27 histone modifications in Arabidopsis, offering three key functionalities. Firstly, QHistone employs machine learning to predict the epigenomic profile of a query protein, characterized by its most associated histone modifications, and uses these modifications to infer the protein's role in transcriptional regulation. Secondly, it predicts synergistic regulatory activities between two proteins by comparing their profiles. Lastly, it detects previously unexplored co-regulating protein pairs by screening all known proteins. QHistone accurately identifies histone modifications associated with specific known proteins, and allows users to computationally validate their results using gene expression data from various plant tissues. These functions demonstrate an useful approach to utilizing epigenome data for gene regulation analysis, making QHistone a valuable resource for the scientific community ( https://qhistone.paoyang.ipmb.sinica.edu.tw )., (© 2024. The Author(s).)

Differential accessibility (DA) analysis of single-cell epigenomics data enables the discovery of regulatory programs that establish cell type identity and steer responses to physiological and pathophysiological perturbations. While many statistical methods to identify DA regions have been developed, the principles that determine the performance of these methods remain unclear. As a result, there is no consensus on the most appropriate statistical methods for DA analysis of single-cell epigenomics data. Here, we present a systematic evaluation of statistical methods that have been applied to identify DA regions in single-cell ATAC-seq (scATAC-seq) data. We leverage a compendium of scATAC-seq experiments with matching bulk ATAC-seq or scRNA-seq in order to assess the accuracy, bias, robustness, and scalability of each statistical method. The structure of our experiments also provides the opportunity to define best practices for the analysis of scATAC-seq data beyond DA itself. We leverage this understanding to develop an R package implementing these best practices., (© 2024. The Author(s).)

Cis-regulatory elements (CREs) are pivotal in orchestrating gene expression throughout diverse biological systems. Accurate identification and in-depth characterization of functional CREs are crucial for decoding gene regulation networks during cellular processes. In this study, we develop Kethoxal-Assisted Single-stranded DNA Assay for Transposase-Accessible Chromatin with Sequencing (KAS-ATAC-seq) to quantitatively analyze the transcriptional activity of CREs. A main advantage of KAS-ATAC-seq lies in its precise measurement of ssDNA levels within both proximal and distal ATAC-seq peaks, enabling the identification of transcriptional regulatory sequences. This feature is particularly adept at defining Single-Stranded Transcribing Enhancers (SSTEs). SSTEs are highly enriched with nascent RNAs and specific transcription factors (TFs) binding sites that define cellular identity. Moreover, KAS-ATAC-seq provides a detailed characterization and functional implications of various SSTE subtypes. Our analysis of CREs during mouse neural differentiation demonstrates that KAS-ATAC-seq can effectively identify immediate-early activated CREs in response to retinoic acid (RA) treatment. Our findings indicate that KAS-ATAC-seq provides more precise annotation of functional CREs in transcription. Future applications of KAS-ATAC-seq would help elucidate the intricate dynamics of gene regulation in diverse biological processes., (© 2024. The Author(s).)

Single-cell assay for transposase-accessible chromatin using sequencing (scATAC-seq) is being increasingly used to study gene regulation. However, major analytical gaps limit its utility in studying gene regulatory programs in complex diseases. In response, MOCHA (Model-based single cell Open CHromatin Analysis) presents major advances over existing analysis tools, including: 1) improving identification of sample-specific open chromatin, 2) statistical modeling of technical drop-out with zero-inflated methods, 3) mitigation of false positives in single cell analysis, 4) identification of alternative transcription-starting-site regulation, and 5) modules for inferring temporal gene regulatory networks from longitudinal data. These advances, in addition to open chromatin analyses, provide a robust framework after quality control and cell labeling to study gene regulatory programs in human disease. We benchmark MOCHA with four state-of-the-art tools to demonstrate its advances. We also construct cross-sectional and longitudinal gene regulatory networks, identifying potential mechanisms of COVID-19 response. MOCHA provides researchers with a robust analytical tool for functional genomic inference from scATAC-seq data., (© 2024. The Author(s).)

Amyotrophic Lateral Sclerosis (ALS), like many other neurodegenerative diseases, is highly heritable, but with only a small fraction of cases explained by monogenic disease alleles. To better understand sporadic ALS, we report epigenomic profiles, as measured by ATAC-seq, of motor neuron cultures derived from a diverse group of 380 ALS patients and 80 healthy controls. We find that chromatin accessibility is heavily influenced by sex, the iPSC cell type of origin, ancestry, and the inherent variance arising from sequencing. Once these covariates are corrected for, we are able to identify ALS-specific signals in the data. Additionally, we find that the ATAC-seq data is able to predict ALS disease progression rates with similar accuracy to methods based on biomarkers and clinical status. These results suggest that iPSC-derived motor neurons recapitulate important disease-relevant epigenomic changes., (© 2024. The Author(s).)

Single-cell multi-omics data integration aims to reduce the omics difference while keeping the cell type difference. However, it is daunting to model and distinguish the two differences due to cell heterogeneity. Namely, even cells of the same omics and type would have various features, making the two differences less significant. In this work, we reveal that instead of being an interference, cell heterogeneity could be exploited to improve data integration. Specifically, we observe that the omics difference varies in cells, and cells with smaller omics differences are easier to be integrated. Hence, unlike most existing works that homogeneously treat and integrate all cells, we propose a multi-omics data integration method (dubbed scBridge) that integrates cells in a heterogeneous manner. In brief, scBridge iterates between i) identifying reliable scATAC-seq cells that have smaller omics differences, and ii) integrating reliable scATAC-seq cells with scRNA-seq data to narrow the omics gap, thus benefiting the integration for the rest cells. Extensive experiments on seven multi-omics datasets demonstrate the superiority of scBridge compared with six representative baselines., (© 2023. Springer Nature Limited.)

Computational cell type identification is a fundamental step in single-cell omics data analysis. Supervised celltyping methods have gained increasing popularity in single-cell RNA-seq data because of the superior performance and the availability of high-quality reference datasets. Recent technological advances in profiling chromatin accessibility at single-cell resolution (scATAC-seq) have brought new insights to the understanding of epigenetic heterogeneity. With continuous accumulation of scATAC-seq datasets, supervised celltyping method specifically designed for scATAC-seq is in urgent need. Here we develop Cellcano, a computational method based on a two-round supervised learning algorithm to identify cell types from scATAC-seq data. The method alleviates the distributional shift between reference and target data and improves the prediction performance. After systematically benchmarking Cellcano on 50 well-designed celltyping tasks from various datasets, we show that Cellcano is accurate, robust, and computationally efficient. Cellcano is well-documented and freely available at https://marvinquiet.github.io/Cellcano/ ., (© 2023. The Author(s).)

In mammals, imprinted genes are regulated by differentially methylated regions (DMRs) that are inherited from germ cells, leading to monoallelic expression in accordance with parent-of-origin. Yet, it is largely unknown how imprinted DMRs are maintained in human embryos despite global DNA demethylation following fertilization. Here, we explored the mechanisms involved in imprinting regulation by employing human parthenogenetic embryonic stem cells (hpESCs), which lack paternal alleles. We show that although global loss of DNA methylation in hpESCs affects most imprinted DMRs, many paternally-expressed genes (PEGs) remain repressed. To search for factors regulating PEGs, we performed a genome-wide CRISPR/Cas9 screen in haploid hpESCs. This revealed ATF7IP as an essential repressor of a set of PEGs, which we further show is also required for silencing sperm-specific genes. Our study reinforces an important role for histone modifications in regulating imprinted genes and suggests a link between parental imprinting and germ cell identity., (© 2021. The Author(s).)

Hepatoblastoma (HB) is the most common pediatric liver malignancy; however, hereditary predisposition and acquired molecular aberrations related to HB clinicopathological diversity are not well understood. Here, we perform an integrative genomic profiling of 163 pediatric liver tumors (154 HBs and nine hepatocellular carcinomas) based on the data acquired from a cohort study (JPLT-2). The total number of somatic mutations is precious low (0.52/Mb on exonic regions) but correlated with age at diagnosis. Telomerase reverse transcriptase (TERT) promoter mutations are prevalent in the tween HBs, selective in the transitional liver cell tumor (TLCT, > 8 years old). DNA methylation profiling reveals that classical HBs are characterized by the specific hypomethylated enhancers, which are enriched with binding sites for ASCL2, a regulatory transcription factor for definitive endoderm in Wnt-pathway. Prolonged upregulation of ASCL2, as well as fetal-liver-like methylation patterns of IGF2 promoters, suggests their "cell of origin" derived from the premature hepatoblast, similar to intestinal epithelial cells, which are highly proliferative. Systematic molecular profiling of HB is a promising approach for understanding the epigenetic drivers of hepatoblast carcinogenesis and deriving clues for risk stratification., (© 2021. The Author(s).)

Chronically elevated intraocular pressure (IOP) is the major risk factor of primary open-angle glaucoma, a leading cause of blindness. Dysfunction of the trabecular meshwork (TM), which controls the outflow of aqueous humor (AqH) from the anterior chamber, is the major cause of elevated IOP. Here, we demonstrate that mice deficient in the Krüppel-like zinc finger transcriptional factor GLI-similar-1 (GLIS1) develop chronically elevated IOP. Magnetic resonance imaging and histopathological analysis reveal that deficiency in GLIS1 expression induces progressive degeneration of the TM, leading to inefficient AqH drainage from the anterior chamber and elevated IOP. Transcriptome and cistrome analyses identified several glaucoma- and extracellular matrix-associated genes as direct transcriptional targets of GLIS1. We also identified a significant association between GLIS1 variant rs941125 and glaucoma in humans (P = 4.73 × 10 -6 ), further supporting a role for GLIS1 into glaucoma etiology. Our study identifies GLIS1 as a critical regulator of TM function and maintenance, AqH dynamics, and IOP., (© 2021. This is a U.S. Government work and not under copyright protection in the US; foreign copyright protection may apply.)

ATRX is a tumor suppressor that has been associated with protection from DNA replication stress, purportedly through resolution of difficult-to-replicate G-quadruplex (G4) DNA structures. While several studies demonstrate that loss of ATRX sensitizes cells to chemical stabilizers of G4 structures, the molecular function of ATRX at G4 regions during replication remains unknown. Here, we demonstrate that ATRX associates with a number of the MCM replication complex subunits and that loss of ATRX leads to G4 structure accumulation at newly synthesized DNA. We show that both the helicase domain of ATRX and its H3.3 chaperone function are required to protect cells from G4-induced replicative stress. Furthermore, these activities are upstream of heterochromatin formation mediated by the histone methyltransferase, ESET, which is the critical molecular event that protects cells from G4-mediated stress. In support, tumors carrying mutations in either ATRX or ESET show increased mutation burden at G4-enriched DNA sequences. Overall, our study provides new insights into mechanisms by which ATRX promotes genome stability with important implications for understanding impacts of its loss on human disease.

Sight depends on the tight cooperation between photoreceptors and pigmented cells, which derive from common progenitors through the bifurcation of a single gene regulatory network into the neural retina (NR) and retinal-pigmented epithelium (RPE) programs. Although genetic studies have identified upstream nodes controlling these networks, their regulatory logic remains poorly investigated. Here, we characterize transcriptome dynamics and chromatin accessibility in segregating NR/RPE populations in zebrafish. We analyze cis-regulatory modules and enriched transcription factor motives to show extensive network redundancy and context-dependent activity. We identify downstream targets, highlighting an early recruitment of desmosomal genes in the flattening RPE and revealing Tead factors as upstream regulators. We investigate the RPE specification network dynamics to uncover an unexpected sequence of transcription factors recruitment, which is conserved in humans. This systematic interrogation of the NR/RPE bifurcation should improve both genetic counseling for eye disorders and hiPSCs-to-RPE differentiation protocols for cell-replacement therapies in degenerative diseases.

In plants, RNA-directed DNA methylation (RdDM) is a well-known de novo DNA methylation pathway that involves two plant-specific RNA polymerases, Pol IV and Pol V. In this study, we discovered and characterized an RdDM factor, RDM15. Through DNA methylome and genome-wide siRNA analyses, we show that RDM15 is required for RdDM-dependent DNA methylation and siRNA accumulation at a subset of RdDM target loci. We show that RDM15 contributes to Pol V-dependent downstream siRNA accumulation and interacts with NRPE3B, a subunit specific to Pol V. We also show that the C-terminal tudor domain of RDM15 specifically recognizes the histone 3 lysine 4 monomethylation (H3K4me1) mark. Structure analysis of RDM15 in complex with the H3K4me1 peptide showed that the RDM15 tudor domain specifically recognizes the monomethyllysine through an aromatic cage and a specific hydrogen bonding network; this chemical feature-based recognition mechanism differs from all previously reported monomethyllysine recognition mechanisms. RDM15 and H3K4me1 have similar genome-wide distribution patterns at RDM15-dependent RdDM target loci, establishing a link between H3K4me1 and RDM15-mediated RdDM in vivo. In summary, we have identified and characterized a histone H3K4me1-specific binding protein as an RdDM component, and structural analysis of RDM15 revealed a chemical feature-based lower methyllysine recognition mechanism.

Hox and ParaHox genes encode transcription factors with similar expression patterns in divergent animals. The Pdx (Xlox) homeobox gene, for example, is expressed in a sharp spatial domain in the endodermal cell layer of the gut in chordates, echinoderms, annelids and molluscs. The significance of comparable gene expression patterns is unclear because it is not known if downstream transcriptional targets are also conserved. Here, we report evidence indicating that a classic transcriptional target of Pdx1 in vertebrates, the insulin gene, is a likely direct target of Pdx in Pacific oyster adults. We show that one insulin-related gene, cgILP, is co-expressed with cgPdx in oyster digestive tissue. Transcriptomic comparison suggests that this tissue plays a similar role to the vertebrate pancreas. Using ATAC-seq and ChIP, we identify an upstream regulatory element of the cgILP gene which shows binding interaction with cgPdx protein in oyster hepatopancreas and demonstrate, using a cell culture assay, that the oyster Pdx can act as a transcriptional activator through this site, possibly in synergy with NeuroD. These data argue that a classic homeodomain-target gene interaction dates back to the origin of Bilateria.

Heterozygous HNF1A gene mutations can cause maturity onset diabetes of the young 3 (MODY3), characterized by insulin secretion defects. However, specific mechanisms of MODY3 in humans remain unclear due to lack of access to diseased human pancreatic cells. Here, we utilize MODY3 patient-derived human induced pluripotent stem cells (hiPSCs) to study the effect(s) of a causal HNF1A +/H126D mutation on pancreatic function. Molecular dynamics simulations predict that the H126D mutation could compromise DNA binding and gene target transcription. Genome-wide RNA-Seq and ChIP-Seq analyses on MODY3 hiPSC-derived endocrine progenitors reveal numerous HNF1A gene targets affected by the mutation. We find decreased glucose transporter GLUT2 expression, which is associated with reduced glucose uptake and ATP production in the MODY3 hiPSC-derived β-like cells. Overall, our findings reveal the importance of HNF1A in regulating GLUT2 and several genes involved in insulin secretion that can account for the insulin secretory defect clinically observed in MODY3 patients.

Changes in the epigenetic regulation of gene expression have a central role in evolution. Here, we extensively profiled a panel of human, chimpanzee, gorilla, orangutan, and macaque lymphoblastoid cell lines (LCLs), using ChIP-seq for five histone marks, ATAC-seq and RNA-seq, further complemented with whole genome sequencing (WGS) and whole genome bisulfite sequencing (WGBS). We annotated regulatory elements (RE) and integrated chromatin contact maps to define gene regulatory architectures, creating the largest catalog of RE in primates to date. We report that epigenetic conservation and its correlation with sequence conservation in primates depends on the activity state of the regulatory element. Our gene regulatory architectures reveal the coordination of different types of components and highlight the role of promoters and intragenic enhancers (gE) in the regulation of gene expression. We observe that most regulatory changes occur in weakly active gE. Remarkably, novel human-specific gE with weak activities are enriched in human-specific nucleotide changes. These elements appear in genes with signals of positive selection and human acceleration, tissue-specific expression, and particular functional enrichments, suggesting that the regulatory evolution of these genes may have contributed to human adaptation.

The ability to target epigenetic marks like DNA methylation to specific loci is important in both basic research and in crop plant engineering. However, heritability of targeted DNA methylation, how it impacts gene expression, and which epigenetic features are required for proper establishment are mostly unknown. Here, we show that targeting the CG-specific methyltransferase M.SssI with an artificial zinc finger protein can establish heritable CG methylation and silencing of a targeted locus in Arabidopsis. In addition, we observe highly heritable widespread ectopic CG methylation mainly over euchromatic regions. This hypermethylation shows little effect on transcription while it triggers a mild but significant reduction in the accumulation of H2A.Z and H3K27me3. Moreover, ectopic methylation occurs preferentially at less open chromatin that lacks positive histone marks. These results outline general principles of the heritability and interaction of CG methylation with other epigenomic features that should help guide future efforts to engineer epigenomes.

Unlike other malignancies, therapeutic options in pancreatic ductal adenocarcinoma (PDAC) are largely limited to cytotoxic chemotherapy without the benefit of molecular markers predicting response. Here we report tumor-cell-intrinsic chromatin accessibility patterns of treatment-naïve surgically resected PDAC tumors that were subsequently treated with (Gem)/Abraxane adjuvant chemotherapy. By ATAC-seq analyses of EpCAM + PDAC malignant epithelial cells sorted from 54 freshly resected human tumors, we show here the discovery of a signature of 1092 chromatin loci displaying differential accessibility between patients with disease free survival (DFS) < 1 year and patients with DFS > 1 year. Analyzing transcription factor (TF) binding motifs within these loci, we identify two TFs (ZKSCAN1 and HNF1b) displaying differential nuclear localization between patients with short vs. long DFS. We further develop a chromatin accessibility microarray methodology termed "ATAC-array", an easy-to-use platform obviating the time and cost of next generation sequencing. Applying this methodology to the original ATAC-seq libraries as well as independent libraries generated from patient-derived organoids, we validate ATAC-array technology in both the original ATAC-seq cohort as well as in an independent validation cohort. We conclude that PDAC prognosis can be predicted by ATAC-array, which represents a low-cost, clinically feasible technology for assessing chromatin accessibility profiles.

ATAC-seq is a widely-applied assay used to measure genome-wide chromatin accessibility; however, its ability to detect active regulatory regions can depend on the depth of sequencing coverage and the signal-to-noise ratio. Here we introduce AtacWorks, a deep learning toolkit to denoise sequencing coverage and identify regulatory peaks at base-pair resolution from low cell count, low-coverage, or low-quality ATAC-seq data. Models trained by AtacWorks can detect peaks from cell types not seen in the training data, and are generalizable across diverse sample preparations and experimental platforms. We demonstrate that AtacWorks enhances the sensitivity of single-cell experiments by producing results on par with those of conventional methods using ~10 times as many cells, and further show that this framework can be adapted to enable cross-modality inference of protein-DNA interactions. Finally, we establish that AtacWorks can enable new biological discoveries by identifying active regulatory regions associated with lineage priming in rare subpopulations of hematopoietic stem cells.

Identification of the cis-regulatory elements controlling cell-type specific gene expression patterns is essential for understanding the origin of cellular diversity. Conventional assays to map regulatory elements via open chromatin analysis of primary tissues is hindered by sample heterogeneity. Single cell analysis of accessible chromatin (scATAC-seq) can overcome this limitation. However, the high-level noise of each single cell profile and the large volume of data pose unique computational challenges. Here, we introduce SnapATAC, a software package for analyzing scATAC-seq datasets. SnapATAC dissects cellular heterogeneity in an unbiased manner and map the trajectories of cellular states. Using the Nyström method, SnapATAC can process data from up to a million cells. Furthermore, SnapATAC incorporates existing tools into a comprehensive package for analyzing single cell ATAC-seq dataset. As demonstration of its utility, SnapATAC is applied to 55,592 single-nucleus ATAC-seq profiles from the mouse secondary motor cortex. The analysis reveals ~370,000 candidate regulatory elements in 31 distinct cell populations in this brain region and inferred candidate cell-type specific transcriptional regulators.

While footprinting analysis of ATAC-seq data can theoretically enable investigation of transcription factor (TF) binding, the lack of a computational tool able to conduct different levels of footprinting analysis has so-far hindered the widespread application of this method. Here we present TOBIAS, a comprehensive, accurate, and fast footprinting framework enabling genome-wide investigation of TF binding dynamics for hundreds of TFs simultaneously. We validate TOBIAS using paired ATAC-seq and ChIP-seq data, and find that TOBIAS outperforms existing methods for bias correction and footprinting. As a proof-of-concept, we illustrate how TOBIAS can unveil complex TF dynamics during zygotic genome activation in both humans and mice, and propose how zygotic Dux activates cascades of TFs, binds to repeat elements and induces expression of novel genetic elements.

Single-cell ATAC-seq (scATAC-seq) profiles the chromatin accessibility landscape at single cell level, thus revealing cell-to-cell variability in gene regulation. However, the high dimensionality and sparsity of scATAC-seq data often complicate the analysis. Here, we introduce a method for analyzing scATAC-seq data, called Single-Cell ATAC-seq analysis via Latent feature Extraction (SCALE). SCALE combines a deep generative framework and a probabilistic Gaussian Mixture Model to learn latent features that accurately characterize scATAC-seq data. We validate SCALE on datasets generated on different platforms with different protocols, and having different overall data qualities. SCALE substantially outperforms the other tools in all aspects of scATAC-seq data analysis, including visualization, clustering, and denoising and imputation. Importantly, SCALE also generates interpretable features that directly link to cell populations, and can potentially reveal batch effects in scATAC-seq experiments.