Search

Genome-wide experiments to map the DNA-binding locations of transcription-associated factors (TFs) have shown that the number of genes bound by a TF far exceeds the number of possible direct target genes. Distinguishing functional from non-functional binding is therefore a major challenge in the study of transcriptional regulation. We hypothesized that functional targets can be discovered by correlating binding and expression profiles across multiple experimental conditions. To test this hypothesis, we obtained ChIP-seq and RNA-seq data from matching cell types from the human ENCODE resource, considered promoter-proximal and distal cumulative regulatory models to map binding sites to genes, and used a combination of linear and non-linear measures to correlate binding and expression data. We found that a high degree of correlation between a gene's TF-binding and expression profiles was significantly more predictive of the gene being differentially expressed upon knockdown of that TF, compared to using binding sites in the cell type of interest only. Remarkably, TF targets predicted from correlation across a compendium of cell types were also predictive of functional targets in other cell types. Finally, correlation across a time course of ChIP-seq and RNA-seq experiments was also predictive of functional TF targets in that tissue., Comment: 15 pages + 8 pages supplementary material; 6 figures, 6 supplementary figures, 5 supplementary tables

Gene regulatory network inference uses genome-wide transcriptome measurements in response to genetic, environmental or dynamic perturbations to predict causal regulatory influences between genes. We hypothesized that evolution also acts as a suitable network perturbation and that integration of data from multiple closely related species can lead to improved reconstruction of gene regulatory networks. To test this hypothesis, we predicted networks from temporal gene expression data for 3,610 genes measured during early embryonic development in six Drosophila species and compared predicted networks to gold standard networks of ChIP-chip and ChIP-seq interactions for developmental transcription factors in five species. We found that (i) the performance of single-species networks was independent of the species where the gold standard was measured; (ii) differences between predicted networks reflected the known phylogeny and differences in biology between the species; (iii) an integrative consensus network which minimized the total number of edge gains and losses with respect to all single-species networks performed better than any individual network. Our results show that in an evolutionarily conserved system, integration of data from comparable experiments in multiple species improves the inference of gene regulatory networks. They provide a basis for future studies on the numerous multi-species gene expression datasets for other biological processes available in the literature., Comment: 10 pages text + 3 figures + 1 table + 2 supplementary figures + 3 supplementary tables

Topological network motifs represent functional relationships within and between regulatory and protein-protein interaction networks. Enriched motifs often aggregate into self-contained units forming functional modules. Theoretical models for network evolution by duplication-divergence mechanisms and for network topology by hierarchical scale-free networks have suggested a one-to-one relation between network motif enrichment and aggregation, but this relation has never been tested quantitatively in real biological interaction networks. Here we introduce a novel method for assessing the statistical significance of network motif aggregation and for identifying clusters of overlapping network motifs. Using an integrated network of transcriptional, posttranslational and protein-protein interactions in yeast we show that network motif aggregation reflects a local modularity property which is independent of network motif enrichment. In particular our method identified novel functional network themes for a set of motifs which are not enriched yet aggregate significantly and challenges the conventional view that network motif enrichment is the most basic organizational principle of complex networks., Comment: 12 pages, 5 figures

We have compared a recently developed module-based algorithm LeMoNe for reverse-engineering transcriptional regulatory networks to a mutual information based direct algorithm CLR, using benchmark expression data and databases of known transcriptional regulatory interactions for Escherichia coli and Saccharomyces cerevisiae. A global comparison using recall versus precision curves hides the topologically distinct nature of the inferred networks and is not informative about the specific subtasks for which each method is most suited. Analysis of the degree distributions and a regulator specific comparison show that CLR is 'regulator-centric', making true predictions for a higher number of regulators, while LeMoNe is 'target-centric', recovering a higher number of known targets for fewer regulators, with limited overlap in the predicted interactions between both methods. Detailed biological examples in E. coli and S. cerevisiae are used to illustrate these differences and to prove that each method is able to infer parts of the network where the other fails. Biological validation of the inferred networks cautions against over-interpreting recall and precision values computed using incomplete reference networks., Comment: 13 pages, 1 table, 6 figures + 6 pages supplementary information (1 table, 5 figures)

"Module networks" are a framework to learn gene regulatory networks from expression data using a probabilistic model in which coregulated genes share the same parameters and conditional distributions. We present a method to infer ensembles of such networks and an averaging procedure to extract the statistically most significant modules and their regulators. We show that the inferred probabilistic models extend beyond the data set used to learn the models., Comment: 5 pages REVTeX, 4 figures

The solution of high-dimensional inference and prediction problems in computational biology is almost always a compromise between mathematical theory and practical constraints such as limited computational resources. As time progresses, computational power increases but well-established inference methods often remain locked in their initial suboptimal solution. We revisit the approach of Segal et al. (2003) to infer regulatory modules and their condition-specific regulators from gene expression data. In contrast to their direct optimization-based solution we use a more representative centroid-like solution extracted from an ensemble of possible statistical models to explain the data. The ensemble method automatically selects a subset of most informative genes and builds a quantitatively better model for them. Genes which cluster together in the majority of models produce functionally more coherent modules. Regulators which are consistently assigned to a module are more often supported by literature, but a single model always contains many regulator assignments not supported by the ensemble. Reliably detecting condition-specific or combinatorial regulation is particularly hard in a single optimum but can be achieved using ensemble averaging., Comment: 8 pages REVTeX, 6 figures

Over the last decade, a large variety of clustering algorithms have been developed to detect coregulatory relationships among genes from microarray gene expression data. Model based clustering approaches have emerged as statistically well grounded methods, but the properties of these algorithms when applied to large-scale data sets are not always well understood. An in-depth analysis can reveal important insights about the performance of the algorithm, the expected quality of the output clusters, and the possibilities for extracting more relevant information out of a particular data set. We have extended an existing algorithm for model based clustering of genes to simultaneously cluster genes and conditions, and used three large compendia of gene expression data for S. cerevisiae to analyze its properties. The algorithm uses a Bayesian approach and a Gibbs sampling procedure to iteratively update the cluster assignment of each gene and condition. For large-scale data sets, the posterior distribution is strongly peaked on a limited number of equiprobable clusterings. A GO annotation analysis shows that these local maxima are all biologically equally significant, and that simultaneously clustering genes and conditions performs better than only clustering genes and assuming independent conditions. A collection of distinct equivalent clusterings can be summarized as a weighted graph on the set of genes, from which we extract fuzzy, overlapping clusters using a graph spectral method. The cores of these fuzzy clusters contain tight sets of strongly coexpressed genes, while the overlaps exhibit relations between genes showing only partial coexpression., Comment: 8 pages, 7 figures

In recent years, several authors have used probabilistic graphical models to learn expression modules and their regulatory programs from gene expression data. Here, we demonstrate the use of the synthetic data generator SynTReN for the purpose of testing and comparing module network learning algorithms. We introduce a software package for learning module networks, called LeMoNe, which incorporates a novel strategy for learning regulatory programs. Novelties include the use of a bottom-up Bayesian hierarchical clustering to construct the regulatory programs, and the use of a conditional entropy measure to assign regulators to the regulation program nodes. Using SynTReN data, we test the performance of LeMoNe in a completely controlled situation and assess the effect of the methodological changes we made with respect to an existing software package, namely Genomica. Additionally, we assess the effect of various parameters, such as the size of the data set and the amount of noise, on the inference performance. Overall, application of Genomica and LeMoNe to simulated data sets gave comparable results. However, LeMoNe offers some advantages, one of them being that the learning process is considerably faster for larger data sets. Additionally, we show that the location of the regulators in the LeMoNe regulation programs and their conditional entropy may be used to prioritize regulators for functional validation, and that the combination of the bottom-up clustering strategy with the conditional entropy-based assignment of regulators improves the handling of missing or hidden regulators., Comment: 13 pages, 6 figures + 2 pages, 2 figures supplementary information