Genome-wide in vivo protein-DNA interactions are routinely mapped using high-throughput chromatin immunoprecipitation (ChIP). ChIP-reported regions are typically investigated for enriched sequence-motifs, which are likely to model the DNA-binding specificity of the profiled protein and/or of co-occurring proteins. However, simple enrichment analyses can miss insights into the binding-activity of the protein. Note that ChIP reports regions making direct contact with the protein as well as those binding through intermediaries. For example, consider a ChIP experiment targeting protein X, which binds DNA at its cognate sites, but simultaneously interacts with four other proteins. Each of these proteins also binds to its own specific cognate sites along distant parts of the genome, a scenario consistent with the current view of transcriptional hubs and chromatin loops. Since ChIP will pull down all X-associated regions, the final reported data will be a union of five distinct sets of regions, each containing binding sites of one of the five proteins, respectively. Characterizing all five different motifs and the corresponding sets is important to interpret the ChIP experiment and ultimately, the role of X in regulation. We present diversity which attempts exactly this: it partitions the data so that each partition can be characterized with its own de novo motif. Diversity uses a Bayesian approach to identify the optimal number of motifs and the associated partitions, which together explain the entire dataset. This is in contrast to standard motif finders, which report motifs individually enriched in the data, but do not necessarily explain all reported regions. We show that the different motifs and associated regions identified by diversity give insights into the various complexes that may be forming along the chromatin, something that has so far not been attempted from ChIP data. Webserver at http://diversity.ncl.res.in/; standalone (Mac OS X/Linux) from https://github.com/NarlikarLab/DIVERSITY/releases/tag/v1.0.0.

Though the sequence of the genome within each eukaryotic cell is essentially fixed, it exists within a complex and changing chromatin state. This state is determined, in part, by the dynamic binding of proteins to the DNA. These proteins—including histones, transcription factors (TFs), and polymerases—interact with one another, the genome, and other molecules to allow the chromatin to adopt one of exceedingly many possible configurations. Understanding how changing chromatin configurations associate with transcription remains a fundamental research problem. We sought to characterize at high spatiotemporal resolution the dynamic interplay between transcription and chromatin in response to cadmium stress. Whereas gene regulatory responses to environmental stress in yeast have been studied, how the chromatin state changes and how those changes connect to gene regulation remain unexplored. By combining MNase-seq and RNA-seq data, we found chromatin signatures of transcriptional activation and repression involving both nucleosomal and TF-sized DNA-binding factors. Using these signatures, we identified associations between chromatin dynamics and transcriptional regulation, not only for known cadmium response genes, but across the entire genome, including antisense transcripts. Those associations allowed us to develop generalizable models that predict dynamic transcriptional responses on the basis of dynamic chromatin signatures.

Summary: Promoters have diverse regulatory architectures and thus activate genes differently. For example, some have a TATA-box, many others do not. Even the ones with it can differ in its position relative to the transcription start site (TSS). No Promoter Left Behind (NPLB) is an efficient, organism-independent method for characterizing such diverse architectures directly from experimentally identified genome-wide TSSs, without relying on known promoter elements. As a test case, we show its application in identifying novel architectures in the fly genome. Availability and implementation: Web-server at http://nplb.ncl.res.in. Standalone also at https://github.com/computationalBiology/NPLB/ (Mac OSX/Linux). Contact: l.narlikar@ncl.res.in Supplementary information: Supplementary data are available at Bioinformatics online.

We present a novel gene-level regulatory model called SCARlink that predicts single-cell gene expression from single-cell chromatin accessibility within and flanking (+/- 250kb) the genic loci by training on multiome (scRNA-seq and scATAC-seq co-assay) sequencing data. The approach uses regularized Poisson regression on tile-level accessibility data to jointly model all regulatory effects at a gene locus, avoiding the limitations of pairwise gene-peak correlations and dependence on a peak atlas. SCARlink significantly outperformed existing gene scoring methods for imputing gene expression from chromatin accessibility across across high-coverage multiome data sets while giving comparable to improved performance on low-coverage data sets. Shapley value analysis on trained models identified cell-type-specific gene enhancers that are validated by promoter capture Hi-C and are 8x-35x enriched in fine-mapped eQTLs and 22x-35x enriched in fine-mapped GWAS variants across 83 UK Biobank traits. We further show that SCARlink-predicted and observed gene expression vectors provide a robust way to compute a chromatin potential vector field to enable developmental trajectory analysis.

AbstractThough the sequence of the genome within each eukaryotic cell is essentially fixed, it exists in a complex and changing chromatin state. This state is determined, in part, by the dynamic binding of proteins to the DNA. These proteins—including histones, transcription factors (TFs), and polymerases—interact with one another, the genome, and other molecules to allow the chromatin to adopt one of exceedingly many possible configurations. Understanding how changing chromatin configurations associate with transcription remains a fundamental research problem. We sought to characterize at high spatiotemporal resolution the dynamic interplay between transcription and chromatin in response to cadmium stress. While gene regulatory responses to environmental stress in yeast have been studied, how the chromatin state is modified and how those modifications connect to gene regulation remain unexplored. By combining MNase-seq and RNA-seq data, we found chromatin signatures of transcriptional activation and repression involving both nucleosomal and TF-sized DNA binding factors. Using these signatures, we identified associations between chromatin dynamics and transcriptional regulation, not only for known cadmium response genes, but across the entire genome, including antisense transcripts. Those associations allowed us to develop generalizable models that can predict dynamic transcriptional responses on the basis of dynamic chromatin signatures.

Chromatin is a tightly packaged structure of DNA and protein within the nucleus of a cell. The arrangement of different protein complexes along the DNA modulates and is modulated by gene expression. Measuring the binding locations and occupancy levels of different transcription factors (TFs) and nucleosomes is therefore crucial to understanding gene regulation. Antibody-based methods for assaying chromatin occupancy are capable of identifying the binding sites of specific DNA binding factors, but only one factor at a time. In contrast, epigenomic accessibility data like MNase-seq, DNase-seq, and ATAC-seq provide insight into the chromatin landscape of all factors bound along the genome, but with little insight into the identities of those factors. Here, we present RoboCOP, a multivariate state space model that integrates chromatin accessibility data with nucleotide sequence to jointly compute genome-wide probabilistic scores of nucleosome and TF occupancy, for hundreds of different factors. We apply RoboCOP to MNase-seq and ATAC-seq data to elucidate the protein-binding landscape of nucleosomes and 150 TFs across the yeast genome, and show that our model makes better predictions than existing methods. We also compute a chromatin occupancy profile of the yeast genome under cadmium stress, revealing chromatin dynamics associated with transcriptional regulation.

1Chromatin is the tightly packaged structure of DNA and protein within the nucleus of a cell. 2The arrangement of different protein complexes along the DNA modulates and is modulated 3 by gene expression. Measuring the binding locations and level of occupancy of different tran-4 scription factors (TFs) and nucleosomes is therefore crucial to understanding gene regulation. 5Antibody-based methods for assaying chromatin occupancy are capable of identifying the bind-6 ing sites of specific DNA binding factors, but only one factor at a time. On the other hand, 7 epigenomic accessibility data like ATAC-seq, DNase-seq, and MNase-seq provide insight into 8 the chromatin landscape of all factors bound along the genome, but with minimal insight into 9 the identities of those factors. Here, we present RoboCOP, a multivariate state space model 10 that integrates chromatin information from epigenomic accessibility data with nucleotide se-11 quence to compute genome-wide probabilistic scores of nucleosome and TF occupancy, for 12 hundreds of different factors at once. We apply RoboCOP to MNase-seq data to elucidate the 13 protein-binding landscape of nucleosomes and 150 TFs across the yeast genome. Using available 14 protein-binding datasets from the literature, we show that our model predicts the binding of 15 these factors genome-wide more accurately than existing methods. 16 1 INTRODUCTION 17 A cell's chromatin consists of the genome and all the proteins and protein complexes arrayed 18 along it. The arrangement of proteins along the genome determines whether and to what 19 extent the cell's various genes are expressed. Therefore, deciphering the chromatin landscape-20 the positions of all the different proteins bound to the DNA-is crucial to developing a more 21 mechanistic and predictive understanding of gene regulation. 22 Two important types of DNA binding factors (DBFs) are transcription factors (TFs) and 23 nucleosomes. TFs are gene regulatory proteins that activate or repress the transcription of 24 genes by binding with specific sequence preferences to sites along the DNA. Nucleosomes form 25 when 147 base pairs of DNA are wrapped around an octamer of histone proteins. They have 26lower sequence specificity than TFs, but still exhibit a preference for a periodic arrangement 27 of dinucleotides that facilitates DNA wrapping (1). Likened to beads on a string, nucleosomes 28 are positioned fairly regularly along the DNA, occupying about 81% of the genome in the case 29 of the yeast Saccharomyces cerevisiae (2). In taking up their respective positions, nucleosomes 30 contribute to the regulation of gene expression in part by allowing or blocking TFs from oc-31 cupying their putative binding sites. Useful models of the chromatin landscape must therefore 32 be able to simultaneously represent and reason about many DBFs at once, and must explicitly 33 account for the way they compete with one another to bind the genome. 34The binding locations of DBFs have been assayed extensively at high resolution with 35 antibody-based...

No abstract