Search

Proper maintenance of epigenetic information after replication is dependent on the rapid assembly and maturation of chromatin. Chromatin Assembly Complex 1 (CAF-1) is a conserved histone chaperone that deposits (H3-H4)2tetramers as part of the replication-dependent chromatin assembly process. Loss of CAF-1 leads to a delay in chromatin maturation, albeit with minimal impact on steady-state chromatin structure. However, the mechanisms by which CAF-1 mediates the deposition of (H3-H4)2tetramers and the phenotypic consequences of CAF-1-associated assembly defects are not well understood. We used nascent chromatin occupancy profiling to track the spatiotemporal kinetics of chromatin maturation in both wild-type (WT) and CAF-1 mutant yeast cells. Our results show that loss of CAF-1 leads to a heterogeneous rate of nucleosome assembly, with some nucleosomes maturing at near WT kinetics and others exhibiting significantly slower maturation kinetics. The slow-to-mature nucleosomes are enriched in intergenic and poorly transcribed regions, suggesting that transcription-dependent assembly mechanisms can reset the slow-to-mature nucleosomes following replication. Nucleosomes with slow maturation kinetics are also associated with poly(dA:dT) sequences, which implies that CAF-1 deposits histones in a manner that counteracts resistance from the inflexible DNA sequence, promoting the formation of histone octamers as well as ordered nucleosome arrays. In addition, we demonstrate that the delay in chromatin maturation is accompanied by a transient and S-phase specific loss of gene silencing and transcriptional regulation, revealing that the DNA replication program can directly shape the chromatin landscape and modulate gene expression through the process of chromatin maturation.

Over a thousand different transcription factors (TFs) bind with varying occupancy across the human genome. Chromatin immunoprecipitation (ChIP) can assay occupancy genome-wide, but only one TF at a time, limiting our ability to comprehensively observe the TF occupancy landscape, let alone quantify how it changes across conditions. We developed TF occupancy profiler (TOP), a Bayesian hierarchical regression framework, to profile genome-wide quantitative occupancy of numerous TFs using data from a single chromatin accessibility experiment (DNase- or ATAC-seq). TOP is supervised, and its hierarchical structure allows it to predict the occupancy of any sequence-specific TF, even those never assayed with ChIP. We used TOP to profile the quantitative occupancy of hundreds of sequence-specific TFs at sites throughout the genome and examined how their occupancies changed in multiple contexts: in approximately 200 human cell types, through 12 h of exposure to different hormones, and across the genetic backgrounds of 70 individuals. TOP enables cost-effective exploration of quantitative changes in the landscape of TF binding.

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.

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. While 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 can predict dynamic transcriptional responses on the basis of dynamic chromatin signatures.

Chromatin is the 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 level of occupancy 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. On the other hand, epigenomic accessibility data like ATAC-seq, DNase-seq, and MNase-seq provide insight into the chromatin landscape of all factors bound along the genome, but with minimal insight into the identities of those factors. Here, we present RoboCOP, a multivariate state space model that integrates chromatin information from epigenomic accessibility data with nucleotide sequence to compute genome-wide probabilistic scores of nucleosome and TF occupancy, for hundreds of different factors at once. RoboCOP can be applied to any epigenomic dataset that provides quantitative insight into chromatin accessibility in any organism, but here we apply it to MNase-seq data to elucidate the protein-binding landscape of nucleosomes and 150 TFs across the yeast genome. Using available protein-binding datasets from the literature, we show that our model more accurately predicts the binding of these factors genome-wide.

Over a thousand different transcription factors (TFs) bind with varying occupancy across the human genome. Chromatin immunoprecipitation (ChIP) can assay occupancy genome-wide, but only one TF at a time, limiting our ability to comprehensively observe the TF occupancy landscape, let alone quantify how it changes across conditions. We developed TOP, a Bayesian hierarchical regression framework, to profile genome-wide quantitative occupancy of numerous TFs using data from a single DNase-seq experiment. TOP is supervised, and its hierarchical structure allows it to predict the occupancy of any sequence-specific TF, even those never assayed with ChIP. We used TOP to profile the quantitative occupancy of nearly 1500 human TF motifs, and examined how their occupancies changed genome-wide in multiple contexts: across 178 cell types, over 12 hours of exposure to different hormones, and across the genetic backgrounds of 70 individuals. TOP enables cost-effective exploration of quantitative changes in the landscape of TF binding.

Chromatin is the 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 level of occupancy 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. On the other hand, epigenomic accessibility data like ATAC-seq, DNase-seq, and MNase-seq provide insight into the chromatin landscape of all factors bound along the genome, but with minimal insight into the identities of those factors. Here, we present RoboCOP, a multivariate state space model that integrates chromatin information from epigenomic accessibility data with nucleotide sequence to compute genome-wide probabilistic scores of nucleosome and TF occupancy, for hundreds of different factors at once. We apply RoboCOP to MNase-seq data to elucidate the protein-binding landscape of nucleosomes and 150 TFs across the yeast genome. Using available protein-binding datasets from the literature, we show that our model predicts the binding of these factors genome-wide more accurately than existing methods.

Glucocorticoids are potent steroid hormones that regulate immunity and metabolism by activating the transcription factor (TF) activity of glucocorticoid receptor (GR). Previous models have proposed that DNA binding motifs and sites of chromatin accessibility predetermine GR binding and activity. However, there are vast excesses of both features relative to the number of GR binding sites. Thus, these features alone are unlikely to account for the specificity of GR binding and activity. To identify genomic and epigenetic contributions to GR binding specificity and the downstream changes resultant from GR binding, we performed hundreds of genome-wide measurements of TF binding, epigenetic state, and gene expression across a 12-h time course of glucocorticoid exposure. We found that glucocorticoid treatment induces GR to bind to nearly all pre-established enhancers within minutes. However, GR binds to only a small fraction of the set of accessible sites that lack enhancer marks. Once GR is bound to enhancers, a combination of enhancer motif composition and interactions between enhancers then determines the strength and persistence of GR binding, which consequently correlates with dramatic shifts in enhancer activation. Over the course of several hours, highly coordinated changes in TF binding and histone modification occupancy occur specifically within enhancers, and these changes correlate with changes in the expression of nearby genes. Following GR binding, changes in the binding of other TFs precede changes in chromatin accessibility, suggesting that other TFs are also sensitive to genomic features beyond that of accessibility.

Tissue-specific gene expression is often thought to arise from spatially restricted transcriptional cascades. However, it is unclear how expression is established at the top of these cascades in the absence of pre-existing specificity. We generated a transcriptional network to explore how transcription factor expression is established in the Arabidopsis thaliana root ground tissue. Regulators of the SHORTROOT-SCARECROW transcriptional cascade were validated in planta. At the top of this cascade, we identified both activators and repressors of SHORTROOT. The aggregate spatial expression of these regulators is not sufficient to predict transcriptional specificity. Instead, modeling, transcriptional reporters, and synthetic promoters support a mechanism whereby expression at the top of the SHORTROOT-SCARECROW cascade is established through opposing activities of activators and repressors.

Single cell genomic techniques promise to yield key insights into the dynamic interplay between gene expression and epigenetic modification. However, the experimental difficulty of performing multiple measurements on the same cell currently limits efforts to combine multiple genomic data sets into a united picture of single cell variation. We show that it is possible to construct cell trajectories, reflecting the changes that occur in a sequential biological process, from single cell ATAC-seq, bisulfite sequencing, and ChIP-seq data. In addition, we present an approach called MATCHER that computationally circumvents the experimental difficulties inherent in performing multiple genomic measurements on a single cell by inferring correspondence between single cell transcriptomic and epigenetic measurements performed on different cells of the same type. MATCHER works by first learning a separate manifold for the trajectory of each kind of genomic data, then aligning the manifolds to infer a shared trajectory in which cells measured using different techniques are directly comparable. Using scM&T-seq data, we confirm that MATCHER accurately predicts true single cell correlations between DNA methylation and gene expression without using known cell correspondence information. We also used MATCHER to infer correlations among gene expression, chromatin accessibility, and histone modifications in single mouse embryonic stem cells. These results reveal the dynamic interplay between epigenetic changes and gene expression underlying the transition from pluripotency to differentiation priming. Our work is a first step toward a united picture of heterogeneous transcriptomic and epigenetic states in single cells.

Single cell experimental techniques reveal transcriptomic and epigenetic heterogeneity among cells, but how these are related is unclear. We present MATCHER, an approach for integrating multiple types of single cell measurements. MATCHER uses manifold alignment to infer single cell multi-omic profiles from transcriptomic and epigenetic measurements performed on different cells of the same type. Using scM&T-seq and sc-GEM data, we confirm that MATCHER accurately predicts true single cell correlations between DNA methylation and gene expression without using known cell correspondences. MATCHER also reveals new insights into the dynamic interplay between the transcriptome and epigenome in single embryonic stem cells and induced pluripotent stem cells. Electronic supplementary material The online version of this article (doi:10.1186/s13059-017-1269-0) contains supplementary material, which is available to authorized users.

Motivation: Transcriptional regulation is directly enacted by the interactions between DNA and many proteins, including transcription factors (TFs), nucleosomes and polymerases. A critical step in deciphering transcriptional regulation is to infer, and eventually predict, the precise locations of these interactions, along with their strength and frequency. While recent datasets yield great insight into these interactions, individual data sources often provide only partial information regarding one aspect of the complete interaction landscape. For example, chromatin immunoprecipitation (ChIP) reveals the binding positions of a protein, but only for one protein at a time. In contrast, nucleases like MNase and DNase can be used to reveal binding positions for many different proteins at once, but cannot easily determine the identities of those proteins. Currently, few statistical frameworks jointly model these different data sources to reveal an accurate, holistic view of the in vivo protein–DNA interaction landscape. Results: Here, we develop a novel statistical framework that integrates different sources of experimental information within a thermodynamic model of competitive binding to jointly learn a holistic view of the in vivo protein–DNA interaction landscape. We show that our framework learns an interaction landscape with increased accuracy, explaining multiple sets of data in accordance with thermodynamic principles of competitive DNA binding. The resulting model of genomic occupancy provides a precise mechanistic vantage point from which to explore the role of protein–DNA interactions in transcriptional regulation. Availability and implementation: The C source code for compete and Python source code for MCMC-based inference are available at http://www.cs.duke.edu/∼amink . Contact: amink@cs.duke.edu Supplementary information: Supplementary data are available at Bioinformatics online.

Modern, computationally-driven, interdisciplinary research demands close collaboration between individuals from many disciplines. Systems biology is one such interdisciplinary field, drawing on computational and experimental expertise to elucidate complex biological systems and processes. In our efforts to better understand the regulation of cell division in budding yeast, the development of a shared, project-specific style of communication became just as much a goal of the collaboration as the development of our models. Finding this common language for the collaboration helped us overcome cultural barriers, understand the nuances of each other’s work, and enhance the precision and interpretability of our models.

The participants in a collaborative interdisciplinary project found that developing a shared, project-specific communication style helped them overcome cultural barriers, understand the nuances of each other's work, and enhance the accuracy, interpretability, and utility of their models.

Motivation: To advance understanding of eukaryotic cell division, it is important to observe the process precisely. To this end, researchers monitor changes in dividing cells as they traverse the cell cycle, with the presence or absence of morphological or genetic markers indicating a cell's position in a particular interval of the cell cycle. A wide variety of marker data is available, including information-rich cellular imaging data. However, few formal statistical methods have been developed to use these valuable data sources in estimating how a population of cells progresses through the cell cycle. Furthermore, existing methods are designed to handle only a single binary marker of cell cycle progression at a time. Consequently, they cannot facilitate comparison of experiments involving different sets of markers. Results: Here, we develop a new sampling model to accommodate an arbitrary number of different binary markers that characterize the progression of a population of dividing cells along a branching process. We engineer a strain of Saccharomyces cerevisiae with fluorescently labeled markers of cell cycle progression, and apply our new model to two image datasets we collected from the strain, as well as an independent dataset of different markers. We use our model to estimate the duration of post-cytokinetic attachment between a S.cerevisiae mother and daughter cell. The Java implementation is fast and extensible, and includes a graphical user interface. Our model provides a powerful and flexible cell cycle analysis tool, suitable to any type or combination of binary markers. Availability: The software is available from: http://www.cs.duke.edu/~amink/software/cloccs/. Contact: michael.mayhew@duke.edu; amink@cs.duke.edu

Motivation: Recent advances in high-throughput experimental techniques have yielded a large amount of data on protein–protein interactions (PPIs). Since these interactions can be organized into networks, and since separate PPI networks can be constructed for different species, a natural research direction is the comparative analysis of such networks across species in order to detect conserved functional modules. This is the task of network alignment. Results: Most conventional network alignment algorithms adopt a node-then-edge-alignment paradigm: they first identify homologous proteins across networks and then consider interactions among them to construct network alignments. In this study, we propose an alternative direct-edge-alignment paradigm. Specifically, instead of explicit identification of homologous proteins, we directly infer plausibly alignable PPIs across species by comparing conservation of their constituent domain interactions. We apply our approach to detect conserved protein complexes in yeast–fly and yeast–worm PPI networks, and show that our approach outperforms two recent approaches in most alignment performance metrics. Availability: Supplementary material and source code can be found at http://www.cs.duke.edu/∼amink/. Contact: xinguo@cs.duke.edu

Imprinted genes are essential in embryonic development, and imprinting dysregulation contributes to human disease. We report two new human imprinted genes: KCNK9 is predominantly expressed in the brain, is a known oncogene, and may be involved in bipolar disorder and epilepsy, while DLGAP2 is a candidate bladder cancer tumor suppressor. Both genes lie on chromosome 8, not previously suspected to contain imprinted genes. We identified these genes, along with 154 others, based on the predictions of multiple classification algorithms using DNA sequence characteristics as features. Our findings demonstrate that DNA sequence characteristics, including recombination hot spots, are sufficient to accurately predict the imprinting status of individual genes in the human genome.

Synchronized populations of cells are often used to study dynamic processes during the cell division cycle. However, the analysis of time series measurements made on synchronized populations is confounded by the fact that populations lose synchrony over time. Time series measurements are thus averages over a population distribution that is broadening over time. Moreover, direct comparison of measurements taken from multiple synchrony experiments is difficult, as the kinetics of progression during the time series are rarely comparable. Here, we present a flexible mathematical model that describes the dynamics of population distributions resulting from synchrony loss over time. The model was developed using S. cerevisiae, but we show that it can be easily adapted to predict distributions in other organisms. We demonstrate that the model reliably fits data collected from populations synchronized by multiple techniques, and can accurately predict cell cycle distributions as measured by other experimental assays. To indicate its broad applicability, we show that the model can be used to compare global periodic transcription data sets from different organisms: S. cerevisiae and S. pombe.

Motivation: An important problem in molecular biology is to identify the locations at which a transcription factor (TF) binds to DNA, given a set of DNA sequences believed to be bound by that TF. In previous work, we showed that information in the DNA sequence of a binding site is sufficient to predict the structural class of the TF that binds it. In particular, this suggests that we can predict which locations in any DNA sequence are more likely to be bound by certain classes of TFs than others. Here, we argue that traditional methods for de novo motif finding can be significantly improved by adopting an informative prior probability that a TF binding site occurs at each sequence location. To demonstrate the utility of such an approach, we present priority, a powerful new de novo motif finding algorithm. Results: Using data from TRANSFAC, we train three classifiers to recognize binding sites of basic leucine zipper, forkhead, and basic helix loop helix TFs. These classifiers are used to equip priority with three class-specific priors, in addition to a default prior to handle TFs of other classes. We apply priority and a number of popular motif finding programs to sets of yeast intergenic regions that are reported by ChIP-chip to be bound by particular TFs. priority identifies motifs the other methods fail to identify, and correctly predicts the structural class of the TF recognizing the identified binding sites. Availability: Supplementary material and code can be found at . Contact: lee@cs.duke.edu, raluca@cs.duke.edu, uwe.ohler@duke.edu, amink@cs.duke.edu.

Motivation: A key goal in molecular biology is to understand the mechanisms by which a cell regulates the transcription of its genes. One important aspect of this transcriptional regulation is the binding of transcription factors (TFs) to their specific cis-regulatory counterparts on the DNA. TFs recognize and bind their DNA counterparts according to the structure of their DNA-binding domains (e.g. zinc finger, leucine zipper, homeodomain). The structure of these domains can be used as a basis for grouping TFs into classes. Although the structure of DNA-binding domains varies widely across TFs generally, the TFs within a particular class bind to DNA in a similar fashion, suggesting the existence of class-specific features in the DNA sequences bound by each class of TFs. Results: In this paper, we apply a sparse Bayesian learning algorithm to identify a small set of class-specific features in the DNA sequences bound by different classes of TFs; the algorithm simultaneously learns a true multi-class classifier that uses these features to predict the DNA-binding domain of the TF that recognizes a particular set of DNA sequences. We train our algorithm on the six largest classes in TRANSFAC, comprising a total of 587 TFs. We learn a six-class classifier for this training set that achieves 87% leave-one-out cross-validation accuracy. We also identify features within cis-regulatory sequences that are highly specific to each class of TF, which has significant implications for how TF binding sites should be modeled for the purpose of motif discovery. Contact: lee@cs.duke.edu; amink@cs.duke.edu

Recently developed methods for learning sparse classifiers are among the state-of-the-art in supervised learning. These methods learn classifiers that incorporate weighted sums of basis functions with sparsity-promoting priors encouraging the weight estimates to be either significantly large or exactly zero. From a learning-theoretic perspective, these methods control the capacity of the learned classifier by minimizing the number of basis functions used, resulting in better generalization. This paper presents three contributions related to learning sparse classifiers. First, we introduce a true multiclass formulation based on multinomial logistic regression. Second, by combining a bound optimization approach with a component-wise update procedure, we derive fast exact algorithms for learning sparse multiclass classifiers that scale favorably in both the number of training samples and the feature dimensionality, making them applicable even to large data sets in high-dimensional feature spaces. To the best of our knowledge, these are the first algorithms to perform exact multinomial logistic regression with a sparsity-promoting prior. Third, we show how nontrivial generalization bounds can be derived for our classifier in the binary case. Experimental results on standard benchmark data sets attest to the accuracy, sparsity, and efficiency of the proposed methods.

Motivation: Network inference algorithms are powerful computational tools for identifying putative causal interactions among variables from observational data. Bayesian network inference algorithms hold particular promise in that they can capture linear, non-linear, combinatorial, stochastic and other types of relationships among variables across multiple levels of biological organization. However, challenges remain when applying these algorithms to limited quantities of experimental data collected from biological systems. Here, we use a simulation approach to make advances in our dynamic Bayesian network (DBN) inference algorithm, especially in the context of limited quantities of biological data. Results: We test a range of scoring metrics and search heuristics to find an effective algorithm configuration for evaluating our methodological advances. We also identify sampling intervals and levels of data discretization that allow the best recovery of the simulated networks. We develop a novel influence score for DBNs that attempts to estimate both the sign (activation or repression) and relative magnitude of interactions among variables. When faced with limited quantities of observational data, combining our influence score with moderate data interpolation reduces a significant portion of false positive interactions in the recovered networks. Together, our advances allow DBN inference algorithms to be more effective in recovering biological networks from experimentally collected data. Availability: Source code and simulated data are available upon request. Supplementary information: http://www.jarvislab.net/Bioinformatics/BNAdvances/

Bayesian network methods are useful for elucidating genetic regulatory networks because they can represent more than pair-wise relationships between variables, are resistant to overfitting, and remain robust in the face of noisy data.

INTRODUCTION Brain activity drives both behavior and regulated gene expression in neurons. Although past studies have identified activity-induced signaling and gene regulation cascades in cultured neurons, much less is known about how activity- dependent transcriptional networks are affected by the variations in cell-type composition, network interconnections, and firing patterns that comprise behaviorally active brain circuits in vivo. Dual mechanism model for behaviorally regulated gene expression diversity. ( Left ) Song brain circuit and zebra finch song motif (transcribed using https://my.scorecloud.com). ( Middle ) Song nucleus–specific (RA, red; Area X, blue) singing- regulated genes (gene A and gene B) in response to neural f ring (yellow). ( Right ) Region-general EATF and region-specific TF only bind to genomic DNA (lines) with region-specific acetylated histone 3 (H3K27ac peaks) and then transcribe their mRNAs (green arrow). RATIONALE We tested the hypothesis that behaviorally regulated gene expression is anatomically and temporally diverse and that the key determinants of this diversity are networks of transcription factors, their genomic binding sites, and epigenetic chromatin states. We analyzed genome-wide, singing-regulated gene expression across time in the four major forebrain regions of the song control system in songbirds, a model of speech production in humans. We then performed a transcription factor motif analysis to identify gene regulatory networks enriched in each song nucleus and measured acetylation of histone 3 at lysine 27 (H3K27ac) to identify chromatin regions that were transcriptionally active in the genomes of song nuclei before and after singing. RESULTS We found that singing was associated with differential regulation of about 10% of all genes in the avian genome that came in several waves across time. Less than 1% of these genes were comparably regulated in all song nuclei tested, and these comprised a core set dominated by immediate-early gene (IEG) transcription factors. By contrast, the vast majority of singing-regulated genes were regulated in only one or a subset of song nuclei, such that each song nucleus had its own dominant subset of genes regulated with defined temporal profiles, controlling a variety of functions. The promoters of many of the singing-regulated genes contained binding motifs for known early-activated transcription factors (EATFs) that become active in response to neural firing, some of which were expressed differentially between song nuclei at baseline. One EATF, calcium-response factor ( CaRF ), was tested with RNA interference knockdown in cultured neurons and found to regulate the predicted genes in response to neural activity, but was also found to modulate their expression even at baseline. More strikingly, we found with H3K27ac analysis that many song nucleus–specific singing-regulated genes did not show increased chromatin regulatory element activity after singing but rather already had primed region-specific regulatory activity before singing began. CONCLUSIONS We propose a dual mechanism for the diversity of behaviorally regulated genes across different brain regions in vivo (see the figure). First, the neural activity associated with singing activates EATFs, and some TFs differentially expressed in brain regions at baseline, to trigger region-specific expression of their target genes. Second, brain region–specific enhancers near activity- regulated genes are waiting in an epigenetically primed state, ready to modulate transcription of general and song nucleus–specific genes at a moment’s notice when the neurons fire. The combination of these two mechanisms underlies a great diversity of behaviorally regulated gene expression, permitting each nucleus to perform its particular function in this complex behavior.

Song-learning birds and humans share independently evolved similarities in brain pathways for vocal learning that are essential for song and speech and are not found in most other species. Comparisons of brain transcriptomes of song-learning birds and humans relative to vocal nonlearners identified convergent gene expression specializations in specific song and speech brain regions of avian vocal learners and humans. The strongest shared profiles relate bird motor and striatal song-learning nuclei, respectively, with human laryngeal motor cortex and parts of the striatum that control speech production and learning. Most of the associated genes function in motor control and brain connectivity. Thus, convergent behavior and neural connectivity for a complex trait are associated with convergent specialized expression of multiple genes.

Due to cell-to-cell variability and asymmetric cell division, cells in a synchronized population lose synchrony over time. As a result, time-series measurements from synchronized cell populations do not reflect the underlying dynamics of cell-cycle processes. Here, we present a branching process deconvolution algorithm that learns a more accurate view of dynamic cell-cycle processes, free from the convolution effects associated with imperfect cell synchronization. Through wavelet-basis regularization, our method sharpens signal without sharpening noise and can remarkably increase both the dynamic range and the temporal resolution of time-series data. Although applicable to any such data, we demonstrate the utility of our method by applying it to a recent cell-cycle transcription time course in the eukaryote Saccharomyces cerevisiae . Our method more sensitively detects cell-cycle–regulated transcription and reveals subtle timing differences that are masked in the original population measurements. Our algorithm also explicitly learns distinct transcription programs for mother and daughter cells, enabling us to identify 82 genes transcribed almost entirely in early G1 in a daughter-specific manner.

Identifying binding sites of transcription factors (TFs) is a key task in deciphering transcriptional regulation. ChIP-based methods are used to survey the genomic locations of a single TF in each experiment. But methods combining DNase digestion data with TF binding specificity information could potentially be used to survey the locations of many TFs in the same experiment, provided such methods permit reasonable levels of sensitivity and specificity. Here, we present a simple such method that outperforms a leading recent method, centipede, marginally in human but dramatically in yeast (average auROC across 20 TFs increases from 74% to 94%). Our method is based on logistic regression and thus benefits from supervision, but we show that partially and completely unsupervised variants perform nearly as well. Because the number of parameters in our method is at least an order of magnitude smaller than CENTIPEDE, we dub it MILLIPEDE.

Tat specific factor 1 (Tat-SF1) interacts with components of both the transcription and splicing machineries and has been classified as a transcription-splicing factor. Although its function as an HIV-1 dependency factor has been investigated, relatively little is known about the cellular functions of Tat-SF1. To identify target genes of Tat-SF1, we utilized a combination of RNAi and exon-specific microarrays. These arrays, which survey genome-wide changes in transcript and individual exon levels, revealed 450 genes with transcript level changes upon Tat-SF1 depletion. Strikingly, 98% of these target genes were down-regulated upon depletion, indicating that Tat-SF1 generally activates gene expression. We also identified 89 genes that showed differential exon level changes after Tat-SF1 depletion. The 89 genes showed evidence of many different types of alternative exon use consistent with the regulation of transcription initiation sites and RNA processing. Minimal overlap between genes with transcript-level and exon-level changes suggests that Tat-SF1 does not functionally couple transcription and splicing. Biological processes significantly enriched with transcript- and exon-level targets include the cell cycle and nucleic acid metabolism; the insulin signaling pathway was enriched with Tat-SF1 transcript-level targets but not exon-level targets. Additionally, a hexamer, ATGCCG, was over-represented in the promoter region of genes showing changes in transcription initiation upon Tat-SF1 depletion. This may represent a novel motif that Tat-SF1 recognizes during transcription. Together, these findings suggest that Tat-SF1 functions independently in transcription and splicing of cellular genes.

Transcriptional regulation is largely enacted by transcription factors (TFs) binding DNA Large numbers of TF binding motifs have been revealed by ChIP-chip experiments followed by computational DNA motif discovery However, the success of motif discovery algorithms has been limited when applied to sequences bound in vivo (such as those identified by ChIP-chip) because the observed TF-DNA interactions are not necessarily direct: some TFs predominantly associate with DNA indirectly through protein partners, while others exhibit both direct and indirect binding. We present the first method for distinguishing between direct and indirect TF-DNA interactions, integrating in vivo TF binding data, in vivo nucleosome occupancy data, and DNA binding motifs from in vitro protein binding microarray experiments When applied to yeast ChIP-chip data, our method reveals that only 48% of the data sets can be readily explained by direct binding of the profiled TF, while 16% can be explained by indirect DNA binding In the remaining 36%, none of the motifs used in our analysis was able to explain the ChIP-chip data, either because the data were too noisy or because the set of motifs was incomplete As more in vitro TF DNA binding motifs become available, our method could be used to build a complete catalog of direct and indirect TF-DNA interactions Our method is not restricted to yeast or to ChIP-chip data, but can be applied in any system for which both in vivo binding data and in vitro DNA binding motifs are available.

The origin recognition complex (ORC) is an essential DNA replication initiation factor conserved in all eukaryotes. In Saccharomyces cerevisiae, ORC binds to specific DNA elements; however, in higher eukaryotes, ORC exhibits little sequence specificity in vitro or in vivo. We investigated the genome-wide distribution of ORC in Drosophila and found that ORC localizes to specific chromosomal locations in the absence of any discernible simple motif. Although no clear sequence motif emerged, we were able to use machine learning approaches to accurately discriminate between ORC-associated sequences and ORC-free sequences based solely on primary sequence. The complex sequence features that define ORC binding sites are highly correlated with nucleosome positioning signals and likely represent a preferred nucleosomal landscape for ORC association. Open chromatin appears to be the underlying feature that is deterministic for ORC binding. ORC-associated sequences are enriched for the histone variant, H3.3, often at transcription start sites, and depleted for bulk nucleosomes. The density of ORC binding along the chromosome is reflected in the time at which a sequence replicates, with early replicating sequences having a high density of ORC binding. Finally, we found a high concordance between sites of ORC binding and cohesin loading, suggesting that, in addition to DNA replication, ORC may be required for the loading of cohesin on DNA in Drosophila.

Hundreds of different factors adorn the eukaryotic genome, binding to it in large number. These DNA binding factors (DBFs) include nucleosomes, transcription factors (TFs), and other proteins and protein complexes, such as the origin recognition complex (ORC). DBFs compete with one another for binding along the genome, yet many current models of genome binding do not consider different types of DBFs together simultaneously. Additionally, binding is a stochastic process that results in a continuum of binding probabilities at any position along the genome, but many current models tend to consider positions as being either binding sites or not. Here, we present a model that allows a multitude of DBFs, each at different concentrations, to compete with one another for binding sites along the genome. The result is an “occupancy profile,” a probabilistic description of the DNA occupancy of each factor at each position. We implement our model efficiently as the software package COMPETE. We demonstrate genome-wide and at specific loci how modeling nucleosome binding alters TF binding, and vice versa, and illustrate how factor concentration influences binding occupancy. Binding cooperativity between nearby TFs arises implicitly via mutual competition with nucleosomes. Our method applies not only to TFs, but also recapitulates known occupancy profiles of a well-studied replication origin with and without ORC binding. Importantly, the sequence preferences our model takes as input are derived from in vitro experiments. This ensures that the calculated occupancy profiles are the result of the forces of competition represented explicitly in our model and the inherent sequence affinities of the constituent DBFs.

Transcription factor (TF) binding site discovery is an important step in understanding transcriptional regulation. Many computational tools have already been developed, but their success in detecting TF motifs is still limited. We believe one of the main reasons for the low accuracy of current methods is that they do not take into account the structural aspects of TF-DNA interaction. We have previously shown that knowledge about the structural class of the TF and information about nucleosome occupancy can be used to improve motif discovery. Here, we demonstrate the benefits of using information about the DNA double-helical stability for motif discovery. We notice that, in general, the energy needed to destabilize the DNA double helix is higher at TF binding sites than at random DNA sites. We use this information to derive informative positional priors that we incorporate into a motif finding algorithm. When applied to yeast ChIP-chip data, the new informative priors improve the performance of the motif finder significantly when compared to priors that do not use the energetic stability information.

As an increasing number of eukaryotic genomes are being sequenced, comparative studies aimed at detecting regulatory elements in intergenic sequences are becoming more prevalent. Most comparative methods for transcription factor (TF) binding site discovery make use of global or local alignments of orthologous regulatory regions to assess whether a particular DNA site is conserved across related organisms, and thus more likely to be functional. Since binding sites are usually short, sometimes degenerate, and often independent of orientation, alignment algorithms may not align them correctly. Here, we present a novel, alignment-free approach for incorporating conservation information into TF motif discovery. We relax the definition of conserved sites: we consider a DNA site within a regulatory region to be conserved in an orthologous sequence if it occurs anywhere in that sequence, irrespective of orientation. We use this definition to derive informative priors over DNA sequence positions, and incorporate these priors into a Gibbs sampling algorithm for motif discovery. Our approach is simple and fast. It does not require sequence alignments, nor the phylogenetic relationships between the orthologous sequences, and yet it is more effective on real biological data than methods that do.

Transcription factor (TF) binding site discovery is an important step in understanding transcriptional regulation. Many computational tools have already been developed, but their success in detecting TF motifs is still limited. We believe one of the main reasons for the low accuracy of current methods is that they do not take into account the structural aspects of TF-DNA interaction. We have previously shown that knowledge about the structural class of the TF and information about nucleosome occupancy can be used to improve motif discovery. Here, we demonstrate the benefits of using information about the DNA double-helical stability for motif discovery. We notice that, in general, the energy needed to destabilize the DNA double helix is higher at TF binding sites than at random DNA sites. We use this information to derive informative positional priors that we incorporate into a motif finding algorithm. When applied to yeast ChIP-chip data, the new informative priors improve the performance of the motif finder significantly when compared to priors that do not use the energetic stability information.

A complete understanding of transcriptional regulatory processes in the cell requires identification of transcription factor binding sites on a genomewide scale. Unfortunately, these binding sites are typically short and degenerate, posing a significant statistical challenge: many more matches to known transcription factor binding sites occur in the genome than are actually functional. Chromatin structure is known to play an important role in guiding transcription factors to those sites that are functional. In particular, it has been shown that active regulatory regions are usually depleted of nucleosomes, thereby enabling transcription factors to bind DNA in those regions [1]. In this paper, we describe a novel algorithm which employs an informative prior over DNA sequence positions based on a discriminative view of nucleosome occupancy; the nucleosome occupancy information comes from a recently published computational model [2]. When a Gibbs sampling algorithm with our informative prior is applied to yeast sequencesets identified by ChIP-chip [3], the correct motif is found in 50% more cases than with an uninformative uniform prior. Moreover, if nucleosome occupancy information is not available, our informative prior reduces to a new kind of prior that can exploit discriminative information in a purely generative setting.

In seeking to find diagnostic biomarkers in proteomic spectra, two significant problems arise. First, not only is there noise in the measured intensity at each m/z value, but there is also noise in the measured m/z value itself. Second, the potential for overfitting is severe: it is easy to find features in the spectra that accurately discriminate disease states but have no biological meaning. We address these problems by developing and testing a series of steps for pre-processing proteomic spectra and extracting putatively meaningful features before presentation to feature selection and classification algorithms. These steps include an HMM-based latent spectrum extraction algorithm for fusing the information from multiple replicate spectra obtained from a single tissue sample, a simple algorithm for baseline correction based on a segmented convex hull, a peak identification and quantification algorithm, and a peak registration algorithm to align peaks from multiple tissue samples into common peak registers. We apply these steps to MALDI spectral data collected from normal and tumor lung tissue samples, and then compare the performance of feature selection with FDR followed by classification with an SVM, versus joint feature selection and classification with Bayesian sparse multinomial logistic regression (SMLR). The SMLR approach outperformed FDR+SVM, but both were effective in achieving good diagnostic accuracy with a small number of features. Some of the selected features have previously been investigated as clinical markers for lung cancer diagnosis; some of the remaining features are excellent candidates for further research.

Determining how information flows along anatomical brain pathways is a fundamental requirement for understanding how animals perceive their environments, learn, and behave. Attempts to reveal such neural information flow have been made using linear computational methods, but neural interactions are known to be nonlinear. Here, we demonstrate that a dynamic Bayesian network (DBN) inference algorithm we originally developed to infer nonlinear transcriptional regulatory networks from gene expression data collected with microarrays is also successful at inferring nonlinear neural information flow networks from electrophysiology data collected with microelectrode arrays. The inferred networks we recover from the songbird auditory pathway are correctly restricted to a subset of known anatomical paths, are consistent with timing of the system, and reveal both the importance of reciprocal feedback in auditory processing and greater information flow to higher-order auditory areas when birds hear natural as opposed to synthetic sounds. A linear method applied to the same data incorrectly produces networks with information flow to non-neural tissue and over paths known not to exist. To our knowledge, this study represents the first biologically validated demonstration of an algorithm to successfully infer neural information flow networks., Synopsis One of the challenges in the area of brain research is to decipher networks describing the flow of information among communicating neurons in the form of electrophysiological signals. These networks are thought to be responsible for perceiving and learning about the environment, as well as producing behavior. Monitoring these networks is limited by the number of electrodes that can be placed in the brain of an awake animal, while inferring and reasoning about these networks is limited by the availability of appropriate computational tools. Here, Smith and Yu and colleagues begin to address these issues by implanting microelectrode arrays in the auditory pathway of freely moving songbirds and by analyzing the data using new computational tools they have designed for deciphering networks. The authors find that a dynamic Bayesian network algorithm they developed to decipher gene regulatory networks from gene expression data effectively infers putative information flow networks in the brain from microelectrode array data. The networks they infer conform to known anatomy and other biological properties of the auditory system and offer new insight into how the auditory system processes natural and synthetic sound. The authors believe that their results represent the first validated study of the inference of information flow networks in the brain.

In seeking to find diagnostic biomarkers in proteomic spectra, two significant problems arise. First, not only is there noise in the measured intensity at each m/z value, but there is also noise in the measured m/z value itself. Second, the potential for overfitting is severe: it is easy to find features in the spectra that accurately discriminate disease states but have no biological meaning. We address these problems by developing and testing a series of steps for pre-processing proteomic spectra and extracting putatively meaningful features before presentation to feature selection and classification algorithms. These steps include an HMM-based latent spectrum extraction algorithm for fusing the information from multiple replicate spectra obtained from a single tissue sample, a simple algorithm for baseline correction based on a segmented convex hull, a peak identification and quantification algorithm, and a peak registration algorithm to align peaks from multiple tissue samples into common peak registers. We apply these steps to MALDI spectral data collected from normal and tumor lung tissue samples, and then compare the performance of feature selection with FDR followed by classification with an SVM, versus joint feature selection and classification with Bayesian sparse multinomial logistic regression (SMLR). The SMLR approach outperformed FDR+SVM, but both were effective in achieving good diagnostic accuracy with a small number of features. Some of the selected features have previously been investigated as clinical markers for lung cancer diagnosis; some of the remaining features are excellent candidates for further research.

Imprinted genes are epigenetically modified genes whose expression is determined according to their parent of origin. They are involved in embryonic development, and imprinting dysregulation is linked to cancer, obesity, diabetes, and behavioral disorders such as autism and bipolar disease. Herein, we train a statistical model based on DNA sequence characteristics that not only identifies potentially imprinted genes, but also predicts the parental allele from which they are expressed. Of 23,788 annotated autosomal mouse genes, our model identifies 600 (2.5%) to be potentially imprinted, 64% of which are predicted to exhibit maternal expression. These predictions allowed for the identification of putative candidate genes for complex conditions where parent-of-origin effects are involved, including Alzheimer disease, autism, bipolar disorder, diabetes, male sexual orientation, obesity, and schizophrenia. We observe that the number, type, and relative orientation of repeated elements flanking a gene are particularly important in predicting whether a gene is imprinted.

We present a method for jointly learning dynamic models of transcriptional regulatory networks from gene expression data and transcription factor binding location data. Models are automatically learned using dynamic Bayesian network inference algorithms; joint learning is accomplished by incorporating evidence from gene expression data through the likelihood, and from transcription factor binding location data through the prior. We propose a new informative structure prior with two advantages. First, the prior incorporates evidence from location data probabilistically, allowing it to be weighed against evidence from expression data. Second, the prior takes on a factorable form that is computationally efficient when learning dynamic regulatory networks. Results obtained from both simulated and experimental data from the yeast cell cycle demonstrate that this joint learning algorithm can recover dynamic regulatory networks from multiple types of data that are more accurate than those recovered from each type of data in isolation.

We present a method for jointly learning dynamic models of transcriptional regulatory networks from gene expression data and transcription factor binding location data. Models are automatically learned using dynamic Bayesian network inference algorithms; joint learning is accomplished by incorporating evidence from gene expression data through the likelihood, and from transcription factor binding location data through the prior. We propose a new informative structure prior with two advantages. First, the prior incorporates evidence from location data probabilistically, allowing it to be weighed against evidence from expression data. Second, the prior takes on a factorable form that is computationally efficient when learning dynamic regulatory networks. Results obtained from both simulated and experimental data from the yeast cell cycle demonstrate that this joint learning algorithm can recover dynamic regulatory networks from multiple types of data that are more accurate than those recovered from each type of data in isolation.

Motivation: Duplication of an organism's entire genome is a rare but spectacular event, enabling the rapid emergence of multiple new gene functions. Over time, the parallel linkage of duplicated genes across chromosomes may be disrupted by reciprocal translocations, while the intra-chromosomal order of genes may be shuffled by inversions and transpositions. Some duplicate genes may evolve unrecognizably or be deleted. As a consequence, the only detectable signature of an ancient duplication event in a modern genome may be the presence of various chromosomal segments containing parallel paralogous genes, with each segment appearing exactly twice in the genome. The problem of reconstructing the linkage structure of an ancestral genome before duplication is known as genome halving with unordered chromosomes. Results: In this paper, we derive a new upper bound on the genome halving distance that is tighter than the best known, and a new lower bound that is almost always tighter than the best known. We also define the notion of genome halving diameter, and obtain both upper and lower bounds for it. Our tighter bounds on genome halving distance yield a new algorithm for reconstructing an ancestral duplicated genome. We create a software package GenomeHalving based on this new algorithm and test it on the yeast genome, identifying a sequence of translocations for halving the yeast genome that is shorter than previously conjectured possible. Availability: GenomeHalving is available upon email request. Contact: py@cs.duke.edu; amink@cs.duke.edu

Recent research has demonstrated quite convincingly that accurate cancer diagnosis can be achieved by constructing classifiers that are designed to compare the gene expression profile of a tissue of unknown cancer status to a database of stored expression profiles from tissues of known cancer status. This paper introduces the JCFO, a novel algorithm that uses a sparse Bayesian approach to jointly identify both the optimal nonlinear classifier for diagnosis and the optimal set of genes on which to base that diagnosis. We show that the diagnostic classification accuracy of the proposed algorithm is superior to a number of current state-of-the-art methods in a full leave-one-out cross-validation study of five widely used benchmark datasets. In addition to its superior classification accuracy, the algorithm is designed to automatically identify a small subset of genes (typically around twenty in our experiments) that are capable of providing complete discriminatory information for diagnosis. Focusing attention on a small subset of genes is useful not only because it produces a classifier with good generalization capacity, but also because this set of genes may provide insights into the mechanisms responsible for the disease itself. A number of the genes identified by the JCFO in our experiments are already in use as clinical markers for cancer diagnosis; some of the remaining genes may be excellent candidates for further clinical investigation. If it is possible to identify a small set of genes that is indeed capable of providing complete discrimination, inexpensive diagnostic assays might be widely deployable in clinical settings.

Recent research has demonstrated quite convincingly that accurate cancer diagnosis can be achieved by constructing classifiers that are designed to compare the gene expression profile of a tissue of unknown cancer status to a database of stored expression profiles from tissues of known cancer status. This paper introduces the JCFO, a novel algorithm that uses a sparse Bayesian approach to jointly identify both the optimal nonlinear classifier for diagnosis and the optimal set of genes on which to base that diagnosis. We show that the diagnostic classification accuracy of the proposed algorithm is superior to a number of current state-of-the-art methods in a full leave-one-out cross-validation study of two widely used benchmark datasets. In addition to its superior classification accuracy, the algorithm is designed to automatically identify a small subset of genes (typically around twenty in our experiments) that are capable of providing complete discriminatory information for diagnosis. Focusing attention on a small subset of genes is not only useful because it produces a classifier with good generalization capacity, but also because this set of genes may provide insights into the mechanisms responsible for the disease itself. A number of the genes identified by the JCFO in our experiments are already in use as clinical markers for cancer diagnosis; some of the remaining genes may be excellent candidates for further clinical investigation. If it is possible to identify a small set of genes that is indeed capable of providing complete discrimination, inexpensive diagnostic assays might be widely deployable in clinical settings.