Gene Regulatory Code Research Articles

Single-cell spatial multi-omics and deep learning dissect enhancer-driven gene regulatory networks in liver zonation.

In the mammalian liver, hepatocytes exhibit diverse metabolic and functional profiles based on their location within the liver lobule. However, it is unclear whether this spatial variation, called zonation, is governed by a well-defined gene regulatory code. Here, using a combination of single-cell multiomics, spatial omics, massively parallel reporter assays and deep learning, we mapped enhancer-gene regulatory networks across mouse liver cell types. We found that zonation affects gene expression and chromatin accessibility in hepatocytes, among other cell types. These states are driven by the repressors TCF7L1 and TBX3, alongside other core hepatocyte transcription factors, such as HNF4A, CEBPA, FOXA1 and ONECUT1. To examine the architecture of the enhancers driving these cell states, we trained a hierarchical deep learning model called DeepLiver. Our study provides a multimodal understanding of the regulatory code underlying hepatocyte identity and their zonation state that can be used to engineer enhancers with specific activity levels and zonation patterns.

DeepRCI: predicting RNA‐chromatin interactions via deep learning with multi‐omics data

Chromatin‐associated RNA (caRNA) acts as a ubiquitous epigenetic layer in eukaryotes, and has been reported to be essential in various biological processes. Here, we propose a highly interpretable computational framework, named DeepRCI, to identify the interactions between various types of RNAs and chromatin. DeepRCI can serve as a useful tool for characterizing RNA‐chromatin interactions and studying the underlying gene regulatory code.BackgroundChromatin‐associated RNA (caRNA) acts as a ubiquitous epigenetic layer in eukaryotes, and has been reported to be essential in various biological processes, including gene transcription, chromatin remodeling and cellular differentiation. Recently, numerous experimental techniques have been developed to characterize genome‐wide RNA‐chromatin interactions to understand their underlying biological functions. However, these experimental methods are generally expensive, time‐consuming, and limited in identifying all potential sites, while most of the existing computational methods are restricted to detecting only specific types of RNAs interacting with chromatin.MethodsHere, we propose a highly interpretable computational framework, named DeepRCI, to identify the interactions between various types of RNAs and chromatin. In this framework, we introduce a novel deep learning component called variformer and integrate multi‐omics data to capture intrinsic genomic features at both RNA and DNA levels.ResultsExtensive experiments demonstrate that DeepRCI can detect RNA‐chromatin interactions more accurately when compared to the state‐of‐the‐art baseline prediction methods. Furthermore, the sequence features extracted by DeepRCI can be well matched to known critical gene regulatory components, indicating that our model can provide useful biological insights into understanding the underlying mechanisms of RNA‐chromatin interactions. In addition, based on the prediction results, we further delineate the relationships between RNA‐chromatin interactions and cellular functions, including gene expression and the modulation of cell states.ConclusionsIn summary, DeepRCI can serve as a useful tool for characterizing RNA‐chromatin interactions and studying the underlying gene regulatory code.

Transcription factor binding site orientation and order are major drivers of gene regulatory activity

The gene regulatory code and grammar remain largely unknown, precluding our ability to link phenotype to genotype in regulatory sequences. Here, using a massively parallel reporter assay (MPRA) of 209,440 sequences, we examine all possible pair and triplet combinations, permutations and orientations of eighteen liver-associated transcription factor binding sites (TFBS). We find that TFBS orientation and order have a major effect on gene regulatory activity. Corroborating these results with genomic analyses, we find clear human promoter TFBS orientation biases and similar TFBS orientation and order transcriptional effects in an MPRA that tested 164,307 liver candidate regulatory elements. Additionally, by adding TFBS orientation to a model that predicts expression from sequence we improve performance by 7.7%. Collectively, our results show that TFBS orientation and order have a significant effect on gene regulatory activity and need to be considered when analyzing the functional effect of variants on the activity of these sequences.

Using CRISPR to understand and manipulate gene regulation.

Understanding how genes are expressed in the correct cell types and at the correct level is a key goal of developmental biology research. Gene regulation has traditionally been approached largely through observational methods, whereas perturbational approaches have lacked precision. CRISPR-Cas9 has begun to transform the study of gene regulation, allowing for precise manipulation of genomic sequences, epigenetic functionalization and gene expression. CRISPR-Cas9 technology has already led to the discovery of new paradigms in gene regulation and, as new CRISPR-based tools and methods continue to be developed, promises to transform our knowledge of the gene regulatory code and our ability to manipulate cell fate. Here, we discuss the current and future application of the emerging CRISPR toolbox toward predicting gene regulatory network behavior, improving stem cell disease modeling, dissecting the epigenetic code, reprogramming cell fate and treating diseases of gene dysregulation.

DNABERT: pre-trained Bidirectional Encoder Representations from Transformers model for DNA-language in genome.

Deciphering the language of non-coding DNA is one of the fundamental problems in genome research. Gene regulatory code is highly complex due to the existence of polysemy and distant semantic relationship, which previous informatics methods often fail to capture especially in data-scarce scenarios. To address this challenge, we developed a novel pre-trained bidirectional encoder representation, named DNABERT, to capture global and transferrable understanding of genomic DNA sequences based on up and downstream nucleotide contexts. We compared DNABERT to the most widely used programs for genome-wide regulatory elements prediction and demonstrate its ease of use, accuracy and efficiency. We show that the single pre-trained transformers model can simultaneously achieve state-of-the-art performance on prediction of promoters, splice sites and transcription factor binding sites, after easy fine-tuning using small task-specific labeled data. Further, DNABERT enables direct visualization of nucleotide-level importance and semantic relationship within input sequences for better interpretability and accurate identification of conserved sequence motifs and functional genetic variant candidates. Finally, we demonstrate that pre-trained DNABERT with human genome can even be readily applied to other organisms with exceptional performance. We anticipate that the pre-trained DNABERT model can be fined tuned to many other sequence analyses tasks. The source code, pretrained and finetuned model for DNABERT are available at GitHub (https://github.com/jerryji1993/DNABERT). Supplementary data are available at Bioinformatics online.

Homotypic Clusters of Transcription Factor Binding Sites in the First Large Intron of AGL24 MADS-Box Transcription Factor Are Recruited in the Enhancement of Floral Expression

The occurrence of homotypic clusters of transcription factor binding sites (HCTs) and their contribution to regulatory evolution is documented in animals, but their manifestation in regulating plant genome expression remained an enigma. To explore the existence and functions of HCTs in highly constrained non-coding sequences of MADS-box transcription factors—generally involved in floral organ identity and phase transition—we employed cis-regulatory cluster finding in silico tools augmented with in vitro assays and site-directed mutagenesis in the STMADS11 superclade of this family. Thousands of transcription factor binding sites (TFBSs) organized into HCTs and multiple-factor binding elements belonging to various families of TFs were identified in the promoter as well as in the intronic region of the selected members of STMADS11 subfamily. A total of 151 HCTs were detectable in the defined promoter in comparison with 144 in the intronic regions. The MADS-domain binding HCTs (overlapping CArG-boxes) constitute a major portion (18%) of these HCTs. A protrusive HCT of 11 TFBSs was identified in the 1st large intron of the AGAMOUS LIKE 24 (AGL24) pertinent to the SVP clade of the STMADS11 subfamily but no such cluster could be detected in the region proximal to the core promoter. These HCTs in the intronic region could bind the SEP4 MADS-domain protein as revealed by in vitro assay. Using HCT cluster swapping of CArG-boxes through the GUS reporter assay, it was possible to empirically validate that a change in the position of intronic MADS-box cluster of AGL24 to the promoter region adjacent to the transcriptional start site (TSS) invoke an enhancement of floral expression. This study unveils the occurrence of HCTs in both the promoter and downstream intronic regions in the selected members of the STMADS11 subfamily of plants. These HCTs play an important role in deciphering the gene regulatory code of MADS-box TFs particularly AGL24.

DNA-Binding Specificities of Human Transcription Factors

Although the proteins that read the gene regulatory code, transcription factors (TFs), have been largely identified, it is not well known which sequences TFs can recognize. We have analyzed the sequence-specific binding of human TFs using high-throughput SELEX and ChIP sequencing. A total of 830 binding profiles were obtained, describing 239 distinctly different binding specificities. The models represent the majority of human TFs, approximately doubling the coverage compared to existing systematic studies. Our results reveal additional specificity determinants for a large number of factors for which a partial specificity was known, including a commonly observed A- or T-rich stretch that flanks the core motifs. Global analysis of the data revealed that homodimer orientation and spacing preferences, and base-stacking interactions, have a larger role in TF-DNA binding than previously appreciated. We further describe a binding model incorporating these features that is required to understand binding of TFs to DNA.

Homotypic clusters of transcription factor binding sites are a key component of human promoters and enhancers

Clustering of multiple transcription factor binding sites (TFBSs) for the same transcription factor (TF) is a common feature of cis-regulatory modules in invertebrate animals, but the occurrence of such homotypic clusters of TFBSs (HCTs) in the human genome has remained largely unknown. To explore whether HCTs are also common in human and other vertebrates, we used known binding motifs for vertebrate TFs and a hidden Markov model-based approach to detect HCTs in the human, mouse, chicken, and fugu genomes, and examined their association with cis-regulatory modules. We found that evolutionarily conserved HCTs occupy nearly 2% of the human genome, with experimental evidence for individual TFs supporting their binding to predicted HCTs. More than half of the promoters of human genes contain HCTs, with a distribution around the transcription start site in agreement with the experimental data from the ENCODE project. In addition, almost half of the 487 experimentally validated developmental enhancers contain them as well--a number more than 25-fold larger than expected by chance. We also found evidence of negative selection acting on TFBSs within HCTs, as the conservation of TFBSs is stronger than the conservation of sequences separating them. The important role of HCTs as components of developmental enhancers is additionally supported by a strong correlation between HCTs and the binding of the enhancer-associated coactivator protein Ep300 (also known as p300). Experimental validation of HCT-containing elements in both zebrafish and mouse suggest that HCTs could be used to predict both the presence of enhancers and their tissue specificity, and are thus a feature that can be effectively used in deciphering the gene regulatory code. In conclusion, our results indicate that HCTs are a pervasive feature of human cis-regulatory modules and suggest that they play an important role in gene regulation in the human and other vertebrate genomes.

Feature context-dependency and complexity-reduction in probability landscapes for integrative genomics

BackgroundThe question of how to integrate heterogeneous sources of biological information into a coherent framework that allows the gene regulatory code in eukaryotes to be systematically investigated is one of the major challenges faced by systems biology. Probability landscapes, which include as reference set the probabilistic representation of the genomic sequence, have been proposed as a possible approach to the systematic discovery and analysis of correlations amongst initially heterogeneous and un-relatable descriptions and genome-wide measurements. Much of the available experimental sequence and genome activity information is de facto, but not necessarily obviously, context dependent. Furthermore, the context dependency of the relevant information is itself dependent on the biological question addressed. It is hence necessary to develop a systematic way of discovering the context-dependency of functional genomics information in a flexible, question-dependent manner.ResultsWe demonstrate here how feature context-dependency can be systematically investigated using probability landscapes. Furthermore, we show how different feature probability profiles can be conditionally collapsed to reduce the computational and formal, mathematical complexity of probability landscapes. Interestingly, the possibility of complexity reduction can be linked directly to the analysis of context-dependency.ConclusionThese two advances in our understanding of the properties of probability landscapes not only simplify subsequent cross-correlation analysis in hypothesis-driven model building and testing, but also provide additional insights into the biological gene regulatory problems studied. Furthermore, insights into the nature of individual features and a classification of features according to their minimal context-dependency are achieved. The formal structure proposed contributes to a concrete and tangible basis for attempting to formulate novel mathematical structures for describing gene regulation in eukaryotes on a genome-wide scale.

Probability landscapes for integrative genomics

BackgroundThe comprehension of the gene regulatory code in eukaryotes is one of the major challenges of systems biology, and is a requirement for the development of novel therapeutic strategies for multifactorial diseases. Its bi-fold degeneration precludes brute force and statistical approaches based on the genomic sequence alone. Rather, recursive integration of systematic, whole-genome experimental data with advanced statistical regulatory sequence predictions needs to be developed. Such experimental approaches as well as the prediction tools are only starting to become available and increasing numbers of genome sequences and empirical sequence annotations are under continual discovery-driven change. Furthermore, given the complexity of the question, a decade(s) long multi-laboratory effort needs to be envisioned. These constraints need to be considered in the creation of a framework that can pave a road to successful comprehension of the gene regulatory code.ResultsWe introduce here a concept for such a framework, based entirely on systematic annotation in terms of probability profiles of genomic sequence using any type of relevant experimental and theoretical information and subsequent cross-correlation analysis in hypothesis-driven model building and testing.ConclusionProbability landscapes, which include as reference set the probabilistic representation of the genomic sequence, can be used efficiently to discover and analyze correlations amongst initially heterogeneous and un-relatable descriptions and genome-wide measurements. Furthermore, this structure is usable as a support for automatically generating and testing hypotheses for alternative gene regulatory grammars and the evaluation of those through statistical analysis of the high-dimensional correlations between genomic sequence, sequence annotations, and experimental data. Finally, this structure provides a concrete and tangible basis for attempting to formulate a mathematical description of gene regulation in eukaryotes on a genome-wide scale.

Functional conservation of Rel binding sites in drosophilid genomes.

Evolutionary constraints on gene regulatory elements are poorly understood: Little is known about how the strength of transcription factor binding correlates with DNA sequence conservation, and whether transcription factor binding sites can evolve rapidly while retaining their function. Here we use the model of the NFKB/Rel-dependent gene regulation in divergent Drosophila species to examine the hypothesis that the functional properties of authentic transcription factor binding sites are under stronger evolutionary constraints than the genomic background. Using molecular modeling we compare tertiary structures of the Drosophila Rel family proteins Dorsal, Dif, and Relish and demonstrate that their DNA-binding and protein dimerization domains undergo distinct rates of evolution. The accumulated amino acid changes, however, are unlikely to affect DNA sequence recognition and affinity. We employ our recently developed microarray-based experimental platform and principal coordinates statistical analysis to quantitatively and systematically profile DNA binding affinities of three Drosophila Rel proteins to 10,368 variants of the NFKB recognition sequences. We then correlate the evolutionary divergence of gene regulatory regions with differences in DNA binding affinities. Genome-wide analyses reveal a significant increase in the number of conserved Rel binding sites in promoters of developmental and immune genes. Significantly, the affinity of Rel proteins to these sites was higher than to less conserved sites and was maintained by the conservation of the DNA binding site sequence (static conservation) or in some cases despite significantly diverged sequences (dynamic conservation). We discuss how two types of conservation may contribute to the stabilization and optimization of a functional gene regulatory code in evolution.