SeqGL Identifies Context-Dependent Binding Signals in Genome-Wide Regulatory Element Maps

Deep Learning for Genomics: From Early Neural Nets to Modern Large Language Models

The data explosion driven by advancements in genomic research, such as high-throughput sequencing techniques, is constantly challenging conventional methods used in genomics. In parallel with the urgent demand for robust algorithms, deep learning has succeeded in various fields such as vision, speech, and text processing. Yet genomics entails unique challenges to deep learning, since we expect a superhuman intelligence that explores beyond our knowledge to interpret the genome from deep learning. A powerful deep learning model should rely on the insightful utilization of task-specific knowledge. In this paper, we briefly discuss the strengths of different deep learning models from a genomic perspective so as to fit each particular task with proper deep learning-based architecture, and we remark on practical considerations of developing deep learning architectures for genomics. We also provide a concise review of deep learning applications in various aspects of genomic research and point out current challenges and potential research directions for future genomics applications. We believe the collaborative use of ever-growing diverse data and the fast iteration of deep learning models will continue to contribute to the future of genomics.

maxATAC: Genome-scale transcription-factor binding prediction from ATAC-seq with deep neural networks

Transcription factors read the genome, fundamentally connecting DNA sequence to gene expression across diverse cell types. Determining how, where, and when TFs bind chromatin will advance our understanding of gene regulatory networks and cellular behavior. The 2017 ENCODE-DREAM in vivo Transcription-Factor Binding Site (TFBS) Prediction Challenge highlighted the value of chromatin accessibility data to TFBS prediction, establishing state-of-the-art methods for TFBS prediction from DNase-seq. However, the more recent Assay-for-Transposase-Accessible-Chromatin (ATAC)-seq has surpassed DNase-seq as the most widely-used chromatin accessibility profiling method. Furthermore, ATAC-seq is the only such technique available at single-cell resolution from standard commercial platforms. While ATAC-seq datasets grow exponentially, suboptimal motif scanning is unfortunately the most common method for TFBS prediction from ATAC-seq. To enable community access to state-of-the-art TFBS prediction from ATAC-seq, we (1) curated an extensive benchmark dataset (127 TFs) for ATAC-seq model training and (2) built "maxATAC", a suite of user-friendly, deep neural network models for genome-wide TFBS prediction from ATAC-seq in any cell type. With models available for 127 human TFs, maxATAC is the largest collection of high-performance TFBS prediction models for ATAC-seq. maxATAC performance extends to primary cells and single-cell ATAC-seq, enabling improved TFBS prediction in vivo. We demonstrate maxATAC's capabilities by identifying TFBS associated with allele-dependent chromatin accessibility at atopic dermatitis genetic risk loci.

BindVAE: Dirichlet variational autoencoders for de novo motif discovery from accessible chromatin

We present a novel unsupervised deep learning approach called BindVAE, based on Dirichlet variational autoencoders, for jointly decoding multiple TF binding signals from open chromatin regions. BindVAE can disentangle an input DNA sequence into distinct latent factors that encode cell-type specific in vivo binding signals for individual TFs, composite patterns for TFs involved in cooperative binding, and genomic context surrounding the binding sites. On the task of retrieving the motifs of expressed TFs in a given cell type, BindVAE is competitive with existing motif discovery approaches.

Dynamic regulatory module networks for inference of cell type-specific transcriptional networks

Changes in transcriptional regulatory networks can significantly alter cell fate. To gain insight into transcriptional dynamics, several studies have profiled bulk multi-omic data sets with parallel transcriptomic and epigenomic measurements at different stages of a developmental process. However, integrating these data to infer cell type-specific regulatory networks is a major challenge. We present dynamic regulatory module networks (DRMNs), a novel approach to infer cell type-specific cis-regulatory networks and their dynamics. DRMN integrates expression, chromatin state, and accessibility to predict cis-regulators of context-specific expression, where context can be cell type, developmental stage, or time point, and uses multitask learning to capture network dynamics across linearly and hierarchically related contexts. We applied DRMNs to study regulatory network dynamics in three developmental processes, each showing different temporal relationships and measuring a different combination of regulatory genomic data sets: cellular reprogramming, liver dedifferentiation, and forward differentiation. DRMN identified known and novel regulators driving cell type-specific expression patterns, showing its broad applicability to examine dynamics of gene regulatory networks from linearly and hierarchically related multi-omic data sets.

Annotating functional effects of non-coding variants in neuropsychiatric cell types by deep transfer learning

Genomewide association studies (GWAS) have identified a large number of loci associated with neuropsychiatric traits, however, understanding the molecular mechanisms underlying these loci remains difficult. To help prioritize causal variants and interpret their functions, computational methods have been developed to predict regulatory effects of non-coding variants. An emerging approach to variant annotation is deep learning models that predict regulatory functions from DNA sequences alone. While such models have been trained on large publicly available dataset such as ENCODE, neuropsychiatric trait-related cell types are under-represented in these datasets, thus there is an urgent need of better tools and resources to annotate variant functions in such cellular contexts. To fill this gap, we collected a large collection of neurodevelopment-related cell/tissue types, and trained deep Convolutional Neural Networks (ResNet) using such data. Furthermore, our model, called MetaChrom, borrows information from public epigenomic consortium to improve the accuracy via transfer learning. We show that MetaChrom is substantially better in predicting experimentally determined chromatin accessibility variants than popular variant annotation tools such as CADD and delta-SVM. By combining GWAS data with MetaChrom predictions, we prioritized 31 SNPs for Schizophrenia, suggesting potential risk genes and the biological contexts where they act. In summary, MetaChrom provides functional annotations of any DNA variants in the neuro-development context and the general method of MetaChrom can also be extended to other disease-related cell or tissue types.

Epitome: predicting epigenetic events in novel cell types with multi-cell deep ensemble learning

The accumulation of large epigenomics data consortiums provides us with the opportunity to extrapolate existing knowledge to new cell types and conditions. We propose Epitome, a deep neural network that learns similarities of chromatin accessibility between well characterized reference cell types and a query cellular context, and copies over signal of transcription factor binding and modification of histones from reference cell types when chromatin profiles are similar to the query. Epitome achieves state-of-the-art accuracy when predicting transcription factor binding sites on novel cellular contexts and can further improve predictions as more epigenetic signals are collected from both reference cell types and the query cellular context of interest.

Predicting TF-DNA Binding Motifs from ChIP-seq Datasets Using the Bag-Based Classifier Combined With a Multi-Fold Learning Scheme

The rapid development of high-throughput sequencing technology provides unique opportunities for studying of transcription factor binding sites, but also brings new computational challenges. Recently, a series of discriminative motif discovery (DMD) methods have been proposed and offer promising solutions for addressing these challenges. However, because of the huge computation cost, most of them have to choose approximate schemes that either sacrifice the accuracy of motif representation or tune motif parameter indirectly. In this paper, we propose a bag-based classifier combined with a multi-fold learning scheme (BCMF) to discover motifs from ChIP-seq datasets. First, BCMF formulates input sequences as a labeled bag naturally. Then, a bag-based classifier, combining with a bag feature extracting strategy, is applied to construct the objective function, and a multi-fold learning scheme is used to solve it. Compared with the existing DMD tools, BCMF features three improvements: 1) Learning position weight matrix (PWM) directly in a continuous space; 2) Proposing to represent a positive bag with a feature fused by its k "most positive" patterns. 3) Applying a more advanced learning scheme. The experimental results on 134 ChIP-seq datasets show that BCMF substantially outperforms existing DMD methods (including DREME, HOMER, XXmotif, motifRG, EDCOD and our previous work).

A machine learning-based framework for modeling transcription elongation

RNA polymerase II (Pol II) generally pauses at certain positions along gene bodies, thereby interrupting the transcription elongation process, which is often coupled with various important biological functions, such as precursor mRNA splicing and gene expression regulation. Characterizing the transcriptional elongation dynamics can thus help us understand many essential biological processes in eukaryotic cells. However, experimentally measuring Pol II elongation rates is generally time and resource consuming. We developed PEPMAN (polymerase II elongation pausing modeling through attention-based deep neural network), a deep learning-based model that accurately predicts Pol II pausing sites based on the native elongating transcript sequencing (NET-seq) data. Through fully taking advantage of the attention mechanism, PEPMAN is able to decipher important sequence features underlying Pol II pausing. More importantly, we demonstrated that the analyses of the PEPMAN-predicted results around various types of alternative splicing sites can provide useful clues into understanding the cotranscriptional splicing events. In addition, associating the PEPMAN prediction results with different epigenetic features can help reveal important factors related to the transcription elongation process. All these results demonstrated that PEPMAN can provide a useful and effective tool for modeling transcription elongation and understanding the related biological factors from available high-throughput sequencing data.

Parsing the Functional Impact of Noncoding Genetic Variants in the Brain Epigenome

The heritability of common psychiatric disorders has motivated global efforts to identify risk-associated genetic variants and elucidate molecular pathways connecting DNA sequence to disease-associated brain dysfunction. The overrepresentation of risk variants among gene regulatory loci instead of protein-coding loci, however, poses a unique challenge in discerning which among the many thousands of variants identified contribute functionally to disease etiology. Defined broadly, psychiatric epigenomics seeks to understand the effects of disease-associated genetic variation on functional readouts of chromatin in an effort to prioritize variants in terms of their impact on gene expression in the brain. Here, we provide an overview of epigenomic mapping in the human brain and highlight findings of particular relevance to psychiatric genetics. Computational methods, including convolutional neuronal networks, and other machine learning approaches hold great promise for elucidating the functional impact of both common and rare genetic variants, thereby refining the epigenomic architecture of psychiatric disorders and enabling integrative analyses of regulatory noncoding variants in the context of large population-level genome and phenome databases.

High-Resolution Mapping of Multiway Enhancer-Promoter Interactions Regulating Pathogen Detection

Eukaryotic gene expression regulation involves thousands of distal regulatory elements. Understanding the quantitative contribution of individual enhancers to gene expression is critical for assessing the role of disease-associated genetic risk variants. Yet, we lack the ability to accurately link genes with their distal regulatory elements. To address this, we used 3D enhancer-promoter (E-P) associations identified using split-pool recognition of interactions by tag extension (SPRITE) to build a predictive model of gene expression. Our model dramatically outperforms models using genomic proximity and can be used to determine the quantitative impact of enhancer loss on gene expression in different genetic backgrounds. We show that genes that form stable E-P hubs have less cell-to-cell variability in gene expression. Finally, we identified transcription factors that regulate stimulation-dependent E-P interactions. Together, our results provide a framework for understanding quantitative contributions of E-P interactions and associated genetic variants to gene expression.

Identification of determinants of differential chromatin accessibility through a massively parallel genome-integrated reporter assay

A key mechanism in cellular regulation is the ability of the transcriptional machinery to physically access DNA. Transcription factors interact with DNA to alter the accessibility of chromatin, which enables changes to gene expression during development or disease or as a response to environmental stimuli. However, the regulation of DNA accessibility via the recruitment of transcription factors is difficult to study in the context of the native genome because every genomic site is distinct in multiple ways. Here we introduce the multiplexed integrated accessibility assay (MIAA), an assay that measures chromatin accessibility of synthetic oligonucleotide sequence libraries integrated into a controlled genomic context with low native accessibility. We apply MIAA to measure the effects of sequence motifs on cell type-specific accessibility between mouse embryonic stem cells and embryonic stem cell-derived definitive endoderm cells, screening 7905 distinct DNA sequences. MIAA recapitulates differential accessibility patterns of 100-nt sequences derived from natively differential genomic regions, identifying E-box motifs common to epithelial-mesenchymal transition driver transcription factors in stem cell-specific accessible regions that become repressed in endoderm. We show that a single binding motif for a key regulatory transcription factor is sufficient to open chromatin, and classify sets of stem cell-specific, endoderm-specific, and shared accessibility-modifying transcription factor motifs. We also show that overexpression of two definitive endoderm transcription factors, T and Foxa2, results in changes to accessibility in DNA sequences containing their respective DNA-binding motifs and identify preferential motif arrangements that influence accessibility.

Enhancer Predictions and Genome-Wide Regulatory Circuits

Spatiotemporal control of gene expression during development requires orchestrated activities of numerous enhancers, which are cis-regulatory DNA sequences that, when bound by transcription factors, support selective activation or repression of associated genes. Proper activation of enhancers is critical during embryonic development, adult tissue homeostasis, and regeneration, and inappropriate enhancer activity is often associated with pathological conditions such as cancer. Multiple consortia [e.g., the Encyclopedia of DNA Elements (ENCODE) Consortium and National Institutes of Health Roadmap Epigenomics Mapping Consortium] and independent investigators have mapped putative regulatory regions in a large number of cell types and tissues, but the sequence determinants of cell-specific enhancers are not yet fully understood. Machine learning approaches trained on large sets of these regulatory regions can identify core transcription factor binding sites and generate quantitative predictions of enhancer activity and the impact of sequence variants on activity. Here, we review these computational methods in the context of enhancer prediction and gene regulatory network models specifying cell fate.

Toll-like Receptor (TLR)-induced Rasgef1b expression in macrophages is regulated by NF-κB through its proximal promoter

Ras Guanine Exchange Factor (RasGEF) domain family member 1b is encoded by a Toll-like receptor (TLR)-inducible gene expressed in macrophages, but transcriptional mechanisms that govern its expression are still unknown. Here, we have functionally characterized the 5' flanking Rasgef1b sequence and analyzed its transcriptional activation. We have identified that the inflammation-responsive promoter is contained within a short sequence (-183 to +119) surrounding the transcriptional start site. The promoter sequence is evolutionarily conserved and harbors a cluster of five NF-κB binding sites. Luciferase reporter gene assay showed that the promoter is responsive to TLR activation and RelA or cRel, but not RelB, transcription factors. Besides, site-directed mutagenesis showed that the κB binding sites are required for maximal promoter activation induced by LPS. Analysis by Assay for Transposase-Accessible Chromatin using sequencing (ATAC-seq) revealed that the promoter is located in an accessible chromatin region. More important, Chromatin Immunoprecipitation sequencing (ChIP-seq) showed that RelA is recruited to the promoter region upon LPS stimulation of bone marrow-derived macrophages. Finally, studies with Rela-deficient macrophages or pharmacological inhibition by Bay11-7082 showed that NF-κB is required for optimal Rasgef1b expression induced by TLR agonists. Our data provide evidence of the regulatory mechanism mediated by NF-κB that facilitates Rasgef1b expression after TLR activation in macrophages.

Sequence and chromatin determinants of transcription factor binding and the establishment of cell type-specific binding patterns

Transcription factors (TFs) selectively bind distinct sets of sites in different cell types. Such cell type-specific binding specificity is expected to result from interplay between the TF's intrinsic sequence preferences, cooperative interactions with other regulatory proteins, and cell type-specific chromatin landscapes. Cell type-specific TF binding events are highly correlated with patterns of chromatin accessibility and active histone modifications in the same cell type. However, since concurrent chromatin may itself be a consequence of TF binding, chromatin landscapes measured prior to TF activation provide more useful insights into how cell type-specific TF binding events became established in the first place. Here, we review the various sequence and chromatin determinants of cell type-specific TF binding specificity. We identify the current challenges and opportunities associated with computational approaches to characterizing, imputing, and predicting cell type-specific TF binding patterns. We further focus on studies that characterize TF binding in dynamic regulatory settings, and we discuss how these studies are leading to a more complex and nuanced understanding of dynamic protein-DNA binding activities. We propose that TF binding activities at individual sites can be viewed along a two-dimensional continuum of local sequence and chromatin context. Under this view, cell type-specific TF binding activities may result from either strongly favorable sequence features or strongly favorable chromatin context.

Combining signal and sequence to detect RNA polymerase initiation in ATAC-seq data

The assay for transposase-accessible chromatin followed by sequencing (ATAC-seq) is an inexpensive protocol for measuring open chromatin regions. ATAC-seq is also relatively simple and requires fewer cells than many other high-throughput sequencing protocols. Therefore, it is tractable in numerous settings where other high throughput assays are challenging to impossible. Hence it is important to understand the limits of what can be inferred from ATAC-seq data. In this work, we leverage ATAC-seq to predict the presence of nascent transcription. Nascent transcription assays are the current gold standard for identifying regions of active transcription, including markers for functional transcription factor (TF) binding. We combine mapped short reads from ATAC-seq with the underlying peak sequence, to determine regions of active transcription genome-wide. We show that a hybrid signal/sequence representation classified using recurrent neural networks (RNNs) can identify these regions across different cell types.

CFSP: a collaborative frequent sequence pattern discovery algorithm for nucleic acid sequence classification

BACKGROUND

Conserved nucleic acid sequences play an essential role in transcriptional regulation. The motifs/templates derived from nucleic acid sequence datasets are usually used as biomarkers to predict biochemical properties such as protein binding sites or to identify specific non-coding RNAs. In many cases, template-based nucleic acid sequence classification performs better than some feature extraction methods, such as N-gram and k-spaced pairs classification. The availability of large-scale experimental data provides an unprecedented opportunity to improve motif extraction methods. The process for pattern extraction from large-scale data is crucial for the creation of predictive models.

METHODS

In this article, a Teiresias-like feature extraction algorithm to discover frequent sub-sequences (CFSP) is proposed. Although gaps are allowed in some motif discovery algorithms, the distance and number of gaps are limited. The proposed algorithm can find frequent sequence pairs with a larger gap. The combinations of frequent sub-sequences in given protracted sequences capture the long-distance correlation, which implies a specific molecular biological property. Hence, the proposed algorithm intends to discover the combinations. A set of frequent sub-sequences derived from nucleic acid sequences with order is used as a base frequent sub-sequence array. The mutation information is attached to each sub-sequence array to implement fuzzy matching. Thus, a mutate records a single nucleotide variant or nucleotides insertion/deletion (indel) to encode a slight difference between frequent sequences and a matched subsequence of a sequence under investigation.

CONCLUSIONS

The proposed algorithm has been validated with several nucleic acid sequence prediction case studies. These data demonstrate better results than the recently available feature descriptors based methods based on experimental data sets such as miRNA, piRNA, and Sigma 54 promoters. CFSP is implemented in C++ and shell script; the source code and related data are available at https://github.com/HePeng2016/CFSP.

Prediction of regulatory motifs from human Chip-sequencing data using a deep learning framework

The identification of transcription factor binding sites and cis-regulatory motifs is a frontier whereupon the rules governing protein–DNA binding are being revealed. Here, we developed a new method (DEep Sequence and Shape mOtif or DESSO) for cis-regulatory motif prediction using deep neural networks and the binomial distribution model. DESSO outperformed existing tools, including DeepBind, in predicting motifs in 690 human ENCODE ChIP-sequencing datasets. Furthermore, the deep-learning framework of DESSO expanded motif discovery beyond the state-of-the-art by allowing the identification of known and new protein–protein–DNA tethering interactions in human transcription factors (TFs). Specifically, 61 putative tethering interactions were identified among the 100 TFs expressed in the K562 cell line. In this work, the power of DESSO was further expanded by integrating the detection of DNA shape features. We found that shape information has strong predictive power for TF–DNA binding and provides new putative shape motif information for human TFs. Thus, DESSO improves in the identification and structural analysis of TF binding sites, by integrating the complexities of DNA binding into a deep-learning framework.

Condition-Specific Modeling of Biophysical Parameters Advances Inference of Regulatory Networks

Large-scale inference of eukaryotic transcription-regulatory networks remains challenging. One underlying reason is that existing algorithms typically ignore crucial regulatory mechanisms, such as RNA degradation and post-transcriptional processing. Here, we describe InfereCLaDR, which incorporates such elements and advances prediction in Saccharomyces cerevisiae. First, InfereCLaDR employs a high-quality Gold Standard dataset that we use separately as prior information and for model validation. Second, InfereCLaDR explicitly models transcription factor activity and RNA half-lives. Third, it introduces expression subspaces to derive condition-responsive regulatory networks for every gene. InfereCLaDR’s final network is validated by known data and trends and results in multiple insights. For example, it predicts long half-lives for transcripts of the nucleic acid metabolism genes and members of the cytosolic chaperonin complex as targets of the proteasome regulator Rpn4p. InfereCLaDR demonstrates that more biophysically realistic modeling of regulatory networks advances prediction accuracy both in eukaryotes and prokaryotes.

BindSpace decodes transcription factor binding signals by large-scale sequence embedding

Decoding transcription factor (TF) binding signals in genomic DNA is a fundamental problem. Here we present a prediction model called BindSpace that learns to embed DNA sequences and TF class/family labels into the same space. By training on binding data for hundreds of TFs and embedding over 1M DNA sequences, BindSpace achieves state-of-the-art multiclass binding prediction performance, in vitro and in vivo, and can distinguish signals of closely related TFs.

Building Transcription Factor Binding Site Models to Understand Gene Regulation in Plants

Transcription factors (TFs) are key cellular components that control gene expression. They recognize specific DNA sequences, the TF binding sites (TFBSs), and thus are targeted to specific regions of the genome where they can recruit transcriptional co-factors and/or chromatin regulators to fine-tune spatiotemporal gene regulation. Therefore, the identification of TFBSs in genomic sequences and their subsequent quantitative modeling is of crucial importance for understanding and predicting gene expression. Here, we review how TFBSs can be determined experimentally, how the TFBS models can be constructed in silico, and how they can be optimized by taking into account features such as position interdependence within TFBSs, DNA shape, and/or by introducing state-of-the-art computational algorithms such as deep learning methods. In addition, we discuss the integration of context variables into the TFBS modeling, including nucleosome positioning, chromatin states, methylation patterns, 3D genome architectures, and TF cooperative binding, in order to better predict TF binding under cellular contexts. Finally, we explore the possibilities of combining the optimized TFBS model with technological advances, such as targeted TFBS perturbation by CRISPR, to better understand gene regulation, evolution, and plant diversity.

Epigenomic analysis reveals DNA motifs regulating histone modifications in human and mouse

Histones are modified by enzymes that act in a locus, cell-type, and developmental stage-specific manner. The recruitment of enzymes to chromatin is regulated at multiple levels, including interaction with sequence-specific DNA-binding factors. However, the DNA-binding specificity of the regulatory factors that orchestrate specific histone modifications has not been broadly mapped. We have analyzed 6 histone marks (H3K4me1, H3K4me3, H3K27ac, H3K27me3, K3H9me3, H3K36me3) across 121 human cell types and tissues from the NIH Roadmap Epigenomics Project as well as 8 histone marks (with addition of H3K4me2 and H3K9ac) from the mouse ENCODE Consortium. We have identified 361 and 369 DNA motifs in human and mouse, respectively, that are the most predictive of each histone mark. Interestingly, 107 human motifs are conserved between the two species. In human embryonic cell line H1, we mutated only the found DNA motifs at particular loci and the significant reduction of H3K27ac levels validated the regulatory roles of the perturbed motifs. The functionality of these motifs was also supported by the evidence that histone-associated motifs, especially H3K4me3 motifs, significantly overlap with the expression of quantitative trait loci SNPs in cancer patients more than the known and random motifs. Furthermore, we observed possible feedbacks to control chromatin dynamics as the found motifs appear in the promoters or enhancers associated with various histone modification enzymes. These results pave the way toward revealing the molecular mechanisms of epigenetic events, such as histone modification dynamics and epigenetic priming.

A De Novo Shape Motif Discovery Algorithm Reveals Preferences of Transcription Factors for DNA Shape Beyond Sequence Motifs

DNA shape adds specificity to sequence motifs but has not been explored systematically outside this context. We hypothesized that DNA-binding proteins (DBPs) preferentially occupy DNA with specific structures ("shape motifs") regardless of whether or not these correspond to high information content sequence motifs. We present ShapeMF, a Gibbs sampling algorithm that identifies de novo shape motifs. Using binding data from hundreds of in vivo and in vitro experiments, we show that most DBPs have shape motifs and can occupy these in the absence of sequence motifs. This "shape-only binding" is common for many DBPs and in regions co-bound by multiple DBPs. When shape and sequence motifs co-occur, they can be overlapping, flanking, or separated by consistent spacing. Finally, DBPs within the same protein family have different shape motifs, explaining their distinct genome-wide occupancy despite having similar sequence motifs. These results suggest that shape motifs not only complement sequence motifs but also facilitate recognition of DNA beyond conventionally defined sequence motifs.

DCDE: An Efficient Deep Convolutional Divergence Encoding Method for Human Promoter Recognition

Efficient human promoter feature extraction is still a major challenge in genome analysis as it can better understand human gene regulation and will be useful for experimental guidance. Although many machine learning algorithms have been developed for eukaryotic gene recognition, performance on promoters is unsatisfactory due to the diverse nature. To extract discriminative features from human promoters, an efficient deep convolutional divergence encoding method (DCDE) is proposed based on statistical divergence (SD) and convolutional neural network (CNN). SD can help optimize kmer feature extraction for human promoters. CNN can also be used to automatically extract features in gene analysis. In DCDE, we first perform informative kmers settlement to encode original gene sequences. A series of SD methods can optimize the most discriminative kmers distributions while maintaining important positional information. Then, CNN is utilized to extract lower dimensional deep features by secondary encoding. Finally, we construct a hybrid recognition architecture with multiple support vector machines and a bilayer decision method. It is flexible to add new features or new models and can be extended to identify other genomic functional elements. The extensive experiments demonstrate that DCDE is effective in promoter encoding and can significantly improve the performance of promoter recognition.

Specificity landscapes unmask submaximal binding site preferences of transcription factors

We have developed Differential Specificity and Energy Landscape (DiSEL) analysis to comprehensively compare DNA-protein interactomes (DPIs) obtained by high-throughput experimental platforms and cutting edge computational methods. While high-affinity DNA binding sites are identified by most methods, DiSEL uncovered nuanced sequence preferences displayed by homologous transcription factors. Pairwise analysis of 726 DPIs uncovered homolog-specific differences at moderate- to low-affinity binding sites (submaximal sites). DiSEL analysis of variants of 41 transcription factors revealed that many disease-causing mutations result in allele-specific changes in binding site preferences. We focused on a set of highly homologous factors that have different biological roles but "read" DNA using identical amino acid side chains. Rather than direct readout, our results indicate that DNA noncontacting side chains allosterically contribute to sculpt distinct sequence preferences among closely related members of transcription factor families.

A massively parallel reporter assay reveals context-dependent activity of homeodomain binding sites in vivo

Cone-rod homeobox (CRX) is a paired-like homeodomain transcription factor (TF) and a master regulator of photoreceptor development in vertebrates. The in vitro DNA binding preferences of CRX have been described in detail, but the degree to which in vitro binding affinity is correlated with in vivo enhancer activity is not known. In addition, paired-class homeodomain TFs can bind DNA cooperatively as both homodimers and heterodimers at inverted TAAT half-sites separated by 2 or 3 nucleotides. This dimeric configuration is thought to mediate target specificity, but whether monomeric and dimeric sites encode distinct levels of activity is not known. Here, we used a massively parallel reporter assay to determine how local sequence context shapes the regulatory activity of CRX binding sites in mouse photoreceptors. We assayed inactivating mutations in more than 1700 TF binding sites and found that dimeric CRX binding sites act as stronger enhancers than monomeric CRX binding sites. Furthermore, the activity of dimeric half-sites is cooperative, dependent on a strict 3-bp spacing, and tuned by the identity of the spacer nucleotides. Saturating single-nucleotide mutagenesis of 195 CRX binding sites showed that, on average, changes in TF binding site affinity are correlated with changes in regulatory activity, but this relationship is obscured when considering mutations across multiple cis-regulatory elements (CREs). Taken together, these results demonstrate that the activity of CRX binding sites is highly dependent on sequence context, providing insight into photoreceptor gene regulation and illustrating functional principles of homeodomain binding sites that may be conserved in other cell types.

BROCKMAN: deciphering variance in epigenomic regulators by k-mer factorization

Background

Variation in chromatin organization across single cells can help shed important light on the mechanisms controlling gene expression, but scale, noise, and sparsity pose significant challenges for interpretation of single cell chromatin data. Here, we develop BROCKMAN (Brockman Representation Of Chromatin by K-mers in Mark-Associated Nucleotides), an approach to infer variation in transcription factor (TF) activity across samples through unsupervised analysis of the variation in DNA sequences associated with an epigenomic mark.

Results

BROCKMAN represents each sample as a vector of epigenomic-mark-associated DNA word frequencies, and decomposes the resulting matrix to find hidden structure in the data, followed by unsupervised grouping of samples and identification of the TFs that distinguish groups. Applied to single cell ATAC-seq, BROCKMAN readily distinguished cell types, treatments, batch effects, experimental artifacts, and cycling cells. We show that each variable component in the k-mer landscape reflects a set of co-varying TFs, which are often known to physically interact. For example, in K562 cells, AP-1 TFs were central determinant of variability in chromatin accessibility through their variable expression levels and diverse interactions with other TFs. We provide a theoretical basis for why cooperative TF binding – and any associated epigenomic mark – is inherently more variable than non-cooperative binding.

Conclusions

BROCKMAN and related approaches will help gain a mechanistic understanding of the trans determinants of chromatin variability between cells, treatments, and individuals.

Electronic supplementary material

The online version of this article (10.1186/s12859-018-2255-6) contains supplementary material, which is available to authorized users.

A novel k-mer set memory (KSM) motif representation improves regulatory variant prediction

The representation and discovery of transcription factor (TF) sequence binding specificities is critical for understanding gene regulatory networks and interpreting the impact of disease-associated noncoding genetic variants. We present a novel TF binding motif representation, the k-mer set memory (KSM), which consists of a set of aligned k-mers that are overrepresented at TF binding sites, and a new method called KMAC for de novo discovery of KSMs. We find that KSMs more accurately predict in vivo binding sites than position weight matrix (PWM) models and other more complex motif models across a large set of ChIP-seq experiments. Furthermore, KSMs outperform PWMs and more complex motif models in predicting in vitro binding sites. KMAC also identifies correct motifs in more experiments than five state-of-the-art motif discovery methods. In addition, KSM-derived features outperform both PWM and deep learning model derived sequence features in predicting differential regulatory activities of expression quantitative trait loci (eQTL) alleles. Finally, we have applied KMAC to 1600 ENCODE TF ChIP-seq data sets and created a public resource of KSM and PWM motifs. We expect that the KSM representation and KMAC method will be valuable in characterizing TF binding specificities and in interpreting the effects of noncoding genetic variations.

Genome-wide prediction of cis-regulatory regions using supervised deep learning methods

Background

In the human genome, 98% of DNA sequences are non-protein-coding regions that were previously disregarded as junk DNA. In fact, non-coding regions host a variety of cis-regulatory regions which precisely control the expression of genes. Thus, Identifying active cis-regulatory regions in the human genome is critical for understanding gene regulation and assessing the impact of genetic variation on phenotype. The developments of high-throughput sequencing and machine learning technologies make it possible to predict cis-regulatory regions genome wide.

Results

Based on rich data resources such as the Encyclopedia of DNA Elements (ENCODE) and the Functional Annotation of the Mammalian Genome (FANTOM) projects, we introduce DECRES based on supervised deep learning approaches for the identification of enhancer and promoter regions in the human genome. Due to their ability to discover patterns in large and complex data, the introduction of deep learning methods enables a significant advance in our knowledge of the genomic locations of cis-regulatory regions. Using models for well-characterized cell lines, we identify key experimental features that contribute to the predictive performance. Applying DECRES, we delineate locations of 300,000 candidate enhancers genome wide (6.8% of the genome, of which 40,000 are supported by bidirectional transcription data), and 26,000 candidate promoters (0.6% of the genome).

Conclusion

The predicted annotations of cis-regulatory regions will provide broad utility for genome interpretation from functional genomics to clinical applications. The DECRES model demonstrates potentials of deep learning technologies when combined with high-throughput sequencing data, and inspires the development of other advanced neural network models for further improvement of genome annotations.

Electronic supplementary material

The online version of this article (10.1186/s12859-018-2187-1) contains supplementary material, which is available to authorized users.

Direct AUC optimization of regulatory motifs

Motivation

The discovery of transcription factor binding site (TFBS) motifs is essential for untangling the complex mechanism of genetic variation under different developmental and environmental conditions. Among the huge amount of computational approaches for de novo identification of TFBS motifs, discriminative motif learning (DML) methods have been proven to be promising for harnessing the discovery power of accumulated huge amount of high-throughput binding data. However, they have to sacrifice accuracy for speed and could fail to fully utilize the information of the input sequences.

Results

We propose a novel algorithm called CDAUC for optimizing DML-learned motifs based on the area under the receiver-operating characteristic curve (AUC) criterion, which has been widely used in the literature to evaluate the significance of extracted motifs. We show that when the considered AUC loss function is optimized in a coordinate-wise manner, the cost function of each resultant sub-problem is a piece-wise constant function, whose optimal value can be found exactly and efficiently. Further, a key step of each iteration of CDAUC can be efficiently solved as a computational geometry problem. Experimental results on real world high-throughput datasets illustrate that CDAUC outperforms competing methods for refining DML motifs, while being one order of magnitude faster. Meanwhile, preliminary results also show that CDAUC may also be useful for improving the interpretability of convolutional kernels generated by the emerging deep learning approaches for predicting TF sequences specificities.

Availability and Implementation

CDAUC is available at: https://drive.google.com/drive/folders/0BxOW5MtIZbJjNFpCeHlBVWJHeW8.

Supplementary information

Supplementary data are available at Bioinformatics online.

Chromatin Accessibility-Based Characterization of the Gene Regulatory Network Underlying Plasmodium falciparum Blood-Stage Development

Underlying the development of malaria parasites within erythrocytes and the resulting pathogenicity is a hardwired program that secures proper timing of gene transcription and production of functionally relevant proteins. How stage-specific gene expression is orchestrated in vivo remains unclear. Here, using the assay for transposase accessible chromatin sequencing (ATAC-seq), we identified ∼4,000 regulatory regions in P. falciparum intraerythrocytic stages. The vast majority of these sites are located within 2 kb upstream of transcribed genes and their chromatin accessibility pattern correlates positively with abundance of the respective mRNA transcript. Importantly, these regions are sufficient to drive stage-specific reporter gene expression and DNA motifs enriched in stage-specific sets of regulatory regions interact with members of the P. falciparum AP2 transcription factor family. Collectively, this study provides initial insights into the in vivo gene regulatory network of P. falciparum intraerythrocytic stages and should serve as a valuable resource for future studies.

The cis-regulatory dynamics of embryonic development at single-cell resolution

Understanding how gene regulatory networks control the progressive restriction of cell fates is a long-standing challenge. Recent advances in measuring single cell gene expression are providing new insights into lineage commitment. However, the regulatory events underlying these changes remain elusive. Here we investigate the dynamics of chromatin regulatory landscapes during embryogenesis at single cell resolution. Using single cell combinatorial indexing assay for transposase accessible chromatin (sci-ATAC-seq)1, we profiled chromatin accessibility in over 20,000 single nuclei from fixed Drosophila embryos spanning three landmark embryonic stages: 2-4 hours (hrs) after egg laying (predominantly stage 5 blastoderm nuclei), when each embryo comprises ~6,000 multipotent cells; 6-8hrs (predominantly stage 10-11), to capture a midpoint in embryonic development when major lineages in the mesoderm and ectoderm are specified; and 10-12hrs (predominantly stage 13), when each of the embryo’s >20,000 cells are undergoing terminal differentiation. Our results reveal spatial heterogeneity in the usage of the regulatory genome prior to gastrulation, a feature that aligns with future cell fate, and nuclei can be temporally ordered along developmental trajectories. During mid-embryogenesis, tissue granularity emerges such that individual cell types can be inferred by their chromatin accessibility, while maintaining a signature of their germ layer of origin. The data reveal overlapping usage of regulatory elements between cells of the endoderm and non-myogenic mesoderm, suggesting a common developmental program reminiscent of the mesendoderm lineage in other species2–4. Altogether, we identify over 30,000 distal regulatory elements exhibiting tissue-specific accessibility. We validated the germ layer specificity of a subset of these predicted enhancers in transgenic embryos, achieving 90% accuracy. Overall, our results demonstrate the power of shotgun single cell profiling of embryos to resolve dynamic changes in the chromatin landscape during development, and to uncover the cis-regulatory programs of metazoan germ layers and cell types.

Deconvolving sequence features that discriminate between overlapping regulatory annotations

Genomic loci with regulatory potential can be annotated with various properties. For example, genomic sites bound by a given transcription factor (TF) can be divided according to whether they are proximal or distal to known promoters. Sites can be further labeled according to the cell types and conditions in which they are active. Given such a collection of labeled sites, it is natural to ask what sequence features are associated with each annotation label. However, discovering such label-specific sequence features is often confounded by overlaps between the labels; e.g. if regulatory sites specific to a given cell type are also more likely to be promoter-proximal, it is difficult to assess whether motifs identified in that set of sites are associated with the cell type or associated with promoters. In order to meet this challenge, we developed SeqUnwinder, a principled approach to deconvolving interpretable discriminative sequence features associated with overlapping annotation labels. We demonstrate the novel analysis abilities of SeqUnwinder using three examples. Firstly, SeqUnwinder is able to unravel sequence features associated with the dynamic binding behavior of TFs during motor neuron programming from features associated with chromatin state in the initial embryonic stem cells. Secondly, we characterize distinct sequence properties of multi-condition and cell-specific TF binding sites after controlling for uneven associations with promoter proximity. Finally, we demonstrate the scalability of SeqUnwinder to discover cell-specific sequence features from over one hundred thousand genomic loci that display DNase I hypersensitivity in one or more ENCODE cell lines.

Transcription factor proteins control gene expression by recognizing and interacting with short DNA sequence patterns in regulatory regions on the genome. Current genomics experiments allow us to find regulatory regions associated with a particular biochemical activity over the entire genome; for example, all regions where a particular transcription factor interacts with the genome in a given cell type. Given a collection of regulatory regions, we often aim to discover short DNA sequence patterns that are more common in the collection than in other regions. Performing such “DNA motif-finding” analysis can give us hints about the patterns that determine gene regulation in the analyzed cell type.

Here we describe a new method for DNA motif-finding called SeqUnwinder. Our approach analyzes collections of regulatory regions where each has been labeled according to various biological properties. For example, the labels could correspond to various cell types in which the regulatory region is active. SeqUnwinder then performs machine-learning analysis to unravel DNA sequence features that are characteristic of each label (e.g. features that distinguish regulatory regions in each cell type from other cell types). SeqUnwinder is the first method to enable analysis of regulatory region collections that contain several overlapping labels.

Identification of Human Lineage-Specific Transcriptional Coregulators Enabled by a Glossary of Binding Modules and Tunable Genomic Backgrounds

Transcription factors (TFs) control cellular processes by binding specific DNA motifs to modulate gene expression. Motif enrichment analysis of regulatory regions can identify direct and indirect TF binding sites. Here, we created a glossary of 108 non-redundant TF-8mer "modules" of shared specificity for 671 metazoan TFs from publicly available and new universal protein binding microarray data. Analysis of 239 ENCODE TF chromatin immunoprecipitation sequencing datasets and associated RNA sequencing profiles suggest the 8mer modules are more precise than position weight matrices in identifying indirect binding motifs and their associated tethering TFs. We also developed GENRE (genomically equivalent negative regions), a tunable tool for construction of matched genomic background sequences for analysis of regulatory regions. GENRE outperformed four state-of-the-art approaches to background sequence construction. We used our TF-8mer glossary and GENRE in the analysis of the indirect binding motifs for the co-occurrence of tethering factors, suggesting novel TF-TF interactions. We anticipate that these tools will aid in elucidating tissue-specific gene-regulatory programs.

An NF-κB Transcription-Factor-Dependent Lineage-Specific Transcriptional Program Promotes Regulatory T Cell Identity and Function

Both conventional T (Tconv) cells and regulatory T (Treg) cells are activated through ligation of the T cell receptor (TCR) complex, leading to the induction of the transcription factor NF-κB. In Tconv cells, NF-κB regulates expression of genes essential for T cell activation, proliferation, and function. However the role of NF-κB in Treg function remains unclear. We conditionally deleted canonical NF-κB members p65 and c-Rel in developing and mature Treg cells and found they have unique but partially redundant roles. c-Rel was critical for thymic Treg development while p65 was essential for mature Treg identity and maintenance of immune tolerance. Transcriptome and NF-κB p65 binding analyses demonstrated a lineage specific, NF-κB-dependent transcriptional program, enabled by enhanced chromatin accessibility. These dual roles of canonical NF-κB in Tconv and Treg cells highlight the functional plasticity of the NF-κB signaling pathway and underscores the need for more selective strategies to therapeutically target NF-κB.

Discovery and validation of information theory-based transcription factor and cofactor binding site motifs

Data from ChIP-seq experiments can derive the genome-wide binding specificities of transcription factors (TFs) and other regulatory proteins. We analyzed 765 ENCODE ChIP-seq peak datasets of 207 human TFs with a novel motif discovery pipeline based on recursive, thresholded entropy minimization. This approach, while obviating the need to compensate for skewed nucleotide composition, distinguishes true binding motifs from noise, quantifies the strengths of individual binding sites based on computed affinity and detects adjacent cofactor binding sites that coordinate with the targets of primary, immunoprecipitated TFs. We obtained contiguous and bipartite information theory-based position weight matrices (iPWMs) for 93 sequence-specific TFs, discovered 23 cofactor motifs for 127 TFs and revealed six high-confidence novel motifs. The reliability and accuracy of these iPWMs were determined via four independent validation methods, including the detection of experimentally proven binding sites, explanation of effects of characterized SNPs, comparison with previously published motifs and statistical analyses. We also predict previously unreported TF coregulatory interactions (e.g. TF complexes). These iPWMs constitute a powerful tool for predicting the effects of sequence variants in known binding sites, performing mutation analysis on regulatory SNPs and predicting previously unrecognized binding sites and target genes.

WSMD: weakly-supervised motif discovery in transcription factor ChIP-seq data

Although discriminative motif discovery (DMD) methods are promising for eliciting motifs from high-throughput experimental data, due to consideration of computational expense, most of existing DMD methods have to choose approximate schemes that greatly restrict the search space, leading to significant loss of predictive accuracy. In this paper, we propose Weakly-Supervised Motif Discovery (WSMD) to discover motifs from ChIP-seq datasets. In contrast to the learning strategies adopted by previous DMD methods, WSMD allows a "global" optimization scheme of the motif parameters in continuous space, thereby reducing the information loss of model representation and improving the quality of resultant motifs. Meanwhile, by exploiting the connection between DMD framework and existing weakly supervised learning (WSL) technologies, we also present highly scalable learning strategies for the proposed method. The experimental results on both real ChIP-seq datasets and synthetic datasets show that WSMD substantially outperforms former DMD methods (including DREME, HOMER, XXmotif, motifRG and DECOD) in terms of predictive accuracy, while also achieving a competitive computational speed.

Mocap: large-scale inference of transcription factor binding sites from chromatin accessibility

Differential binding of transcription factors (TFs) at cis-regulatory loci drives the differentiation and function of diverse cellular lineages. Understanding the regulatory interactions that underlie cell fate decisions requires characterizing TF binding sites (TFBS) across multiple cell types and conditions. Techniques, e.g. ChIP-Seq can reveal genome-wide patterns of TF binding, but typically requires laborious and costly experiments for each TF-cell-type (TFCT) condition of interest. Chromosomal accessibility assays can connect accessible chromatin in one cell type to many TFs through sequence motif mapping. Such methods, however, rarely take into account that the genomic context preferred by each factor differs from TF to TF, and from cell type to cell type. To address the differences in TF behaviors, we developed Mocap, a method that integrates chromatin accessibility, motif scores, TF footprints, CpG/GC content, evolutionary conservation and other factors in an ensemble of TFCT-specific classifiers. We show that integration of genomic features, such as CpG islands improves TFBS prediction in some TFCT. Further, we describe a method for mapping new TFCT, for which no ChIP-seq data exists, onto our ensemble of classifiers and show that our cross-sample TFBS prediction method outperforms several previously described methods.

Inference of cell type specific regulatory networks on mammalian lineages

Transcriptional regulatory networks are at the core of establishing cell type specific gene expression programs. In mammalian systems, such regulatory networks are determined by multiple levels of regulation, including by transcription factors, chromatin environment, and three-dimensional organization of the genome. Recent efforts to measure diverse regulatory genomic datasets across multiple cell types and tissues offer unprecedented opportunities to examine the context-specificity and dynamics of regulatory networks at a greater resolution and scale than before. In parallel, numerous computational approaches to analyze these data have emerged that serve as important tools for understanding mammalian cell type specific regulation. In this article, we review recent computational approaches to predict the expression and sequence-based regulators of a gene's expression level and examine long-range gene regulation. We highlight promising approaches, insights gained, and open challenges that need to be overcome to build a comprehensive picture of cell type specific transcriptional regulatory networks.

DEEP MOTIF DASHBOARD: VISUALIZING AND UNDERSTANDING GENOMIC SEQUENCES USING DEEP NEURAL NETWORKS

Deep neural network (DNN) models have recently obtained state-of-the-art prediction accuracy for the transcription factor binding (TFBS) site classification task. However, it remains unclear how these approaches identify meaningful DNA sequence signals and give insights as to why TFs bind to certain locations. In this paper, we propose a toolkit called the Deep Motif Dashboard (DeMo Dashboard) which provides a suite of visualization strategies to extract motifs, or sequence patterns from deep neural network models for TFBS classification. We demonstrate how to visualize and understand three important DNN models: convolutional, recurrent, and convolutional-recurrent networks. Our first visualization method is finding a test sequence's saliency map which uses first-order derivatives to describe the importance of each nucleotide in making the final prediction. Second, considering recurrent models make predictions in a temporal manner (from one end of a TFBS sequence to the other), we introduce temporal output scores, indicating the prediction score of a model over time for a sequential input. Lastly, a class-specific visualization strategy finds the optimal input sequence for a given TFBS positive class via stochastic gradient optimization. Our experimental results indicate that a convolutional-recurrent architecture performs the best among the three architectures. The visualization techniques indicate that CNN-RNN makes predictions by modeling both motifs as well as dependencies among them.

A synergistic DNA logic predicts genome-wide chromatin accessibility

Enhancers and promoters commonly occur in accessible chromatin characterized by depleted nucleosome contact; however, it is unclear how chromatin accessibility is governed. We show that log-additive cis-acting DNA sequence features can predict chromatin accessibility at high spatial resolution. We develop a new type of high-dimensional machine learning model, the Synergistic Chromatin Model (SCM), which when trained with DNase-seq data for a cell type is capable of predicting expected read counts of genome-wide chromatin accessibility at every base from DNA sequence alone, with the highest accuracy at hypersensitive sites shared across cell types. We confirm that a SCM accurately predicts chromatin accessibility for thousands of synthetic DNA sequences using a novel CRISPR-based method of highly efficient site-specific DNA library integration. SCMs are directly interpretable and reveal that a logic based on local, nonspecific synergistic effects, largely among pioneer TFs, is sufficient to predict a large fraction of cellular chromatin accessibility in a wide variety of cell types.

Basset: learning the regulatory code of the accessible genome with deep convolutional neural networks

The complex language of eukaryotic gene expression remains incompletely understood. Despite the importance suggested by many noncoding variants statistically associated with human disease, nearly all such variants have unknown mechanisms. Here, we address this challenge using an approach based on a recent machine learning advance-deep convolutional neural networks (CNNs). We introduce the open source package Basset to apply CNNs to learn the functional activity of DNA sequences from genomics data. We trained Basset on a compendium of accessible genomic sites mapped in 164 cell types by DNase-seq, and demonstrate greater predictive accuracy than previous methods. Basset predictions for the change in accessibility between variant alleles were far greater for Genome-wide association study (GWAS) SNPs that are likely to be causal relative to nearby SNPs in linkage disequilibrium with them. With Basset, a researcher can perform a single sequencing assay in their cell type of interest and simultaneously learn that cell's chromatin accessibility code and annotate every mutation in the genome with its influence on present accessibility and latent potential for accessibility. Thus, Basset offers a powerful computational approach to annotate and interpret the noncoding genome.

LS-GKM: a new gkm-SVM for large-scale datasets

gkm-SVM is a sequence-based method for predicting and detecting the regulatory vocabulary encoded in functional DNA elements, and is a commonly used tool for studying gene regulatory mechanisms. Here we introduce new software, LS-GKM, which removes several limitations of our previous releases, enabling training on much larger scale (LS) datasets. LS-GKM also provides additional advanced gapped k-mer based kernel functions. With these improvements, LS-GKM achieves considerably higher accuracy than the original gkm-SVM.

AVAILABILITY AND IMPLEMENTATION

C/C ++ source codes and related scripts are freely available from http://github.com/Dongwon-Lee/lsgkm/, and supported on Linux and Mac OS X.

CONTACT

dwlee@jhu.edu

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Early enhancer establishment and regulatory locus complexity shape transcriptional programs in hematopoietic differentiation

We carried out an integrative analysis of enhancer landscape and gene expression dynamics during hematopoietic differentiation using DNase-seq, histone mark ChIP-seq and RNA sequencing to model how the early establishment of enhancers and regulatory locus complexity govern gene expression changes at cell state transitions. We found that high-complexity genes-those with a large total number of DNase-mapped enhancers across the lineage-differ architecturally and functionally from low-complexity genes, achieve larger expression changes and are enriched for both cell type-specific and transition enhancers, which are established in hematopoietic stem and progenitor cells and maintained in one differentiated cell fate but lost in others. We then developed a quantitative model to accurately predict gene expression changes from the DNA sequence content and lineage history of active enhancers. Our method suggests a new mechanistic role for PU.1 at transition peaks during B cell specification and can be used to correct assignments of enhancers to genes.