Dataset Information

Constructing tissue-specific transcriptional regulatory networks via a Markov random field.

ABSTRACT: BACKGROUND:Recent advances in sequencing technologies have enabled parallel assays of chromatin accessibility and gene expression for major human cell lines. Such innovation provides a great opportunity to decode phenotypic consequences of genetic variation via the construction of predictive gene regulatory network models. However, there still lacks a computational method to systematically integrate chromatin accessibility information with gene expression data to recover complicated regulatory relationships between genes in a tissue-specific manner. RESULTS:We propose a Markov random field (MRF) model for constructing tissue-specific transcriptional regulatory networks via integrative analysis of DNase-seq and RNA-seq data. Our method, named CSNets (cell-line specific regulatory networks), first infers regulatory networks for individual cell lines using chromatin accessibility information, and then fine-tunes these networks using the MRF based on pairwise similarity between cell lines derived from gene expression data. Using this method, we constructed regulatory networks specific to 110 human cell lines and 13 major tissues with the use of ENCODE data. We demonstrated the high quality of these networks via comprehensive statistical analysis based on ChIP-seq profiles, functional annotations, taxonomic analysis, and literature surveys. We further applied these networks to analyze GWAS data of Crohn's disease and prostate cancer. Results were either consistent with the literature or provided biological insights into regulatory mechanisms of these two complex diseases. The website of CSNets is freely available at http://bioinfo.au.tsinghua.edu.cn/jianglab/CSNETS/ . CONCLUSIONS:CSNets demonstrated the power of joint analysis on epigenomic and transcriptomic data towards the accurate construction of gene regulatory network. Our work provides not only a useful resource of regulatory networks to the community, but also valuable experiences in methodology development for multi-omics data integration.

SUBMITTER: Ma S

PROVIDER: S-EPMC6311931 | biostudies-literature | 2018 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Constructing tissue-specific transcriptional regulatory networks via a Markov random field.

Ma Shining S Jiang Tao T Jiang Rui R

BMC genomics 20181231 Suppl 10

<h4>Background</h4>Recent advances in sequencing technologies have enabled parallel assays of chromatin accessibility and gene expression for major human cell lines. Such innovation provides a great opportunity to decode phenotypic consequences of genetic variation via the construction of predictive gene regulatory network models. However, there still lacks a computational method to systematically integrate chromatin accessibility information with gene expression data to recover complicated regu ...[more]

PMID: 30598101

Similar Datasets

Project description:BACKGROUND:Transcription factors (TFs) are proteins that can bind to DNA sequences and regulate gene expression. Many TFs are master regulators in cells that contribute to tissue-specific and cell-type-specific gene expression patterns in eukaryotes. Maize has been a model organism for over one hundred years, but little is known about its tissue-specific gene regulation through TFs. In this study, we used a network approach to elucidate gene regulatory networks (GRNs) in four tissues (leaf, root, SAM and seed) in maize. We utilized GENIE3, a machine-learning algorithm combined with large quantity of RNA-Seq expression data to construct four tissue-specific GRNs. Unlike some other techniques, this approach is not limited by high-quality Position Weighed Matrix (PWM), and can therefore predict GRNs for over 2000 TFs in maize. RESULTS:Although many TFs were expressed across multiple tissues, a multi-tiered analysis predicted tissue-specific regulatory functions for many transcription factors. Some well-studied TFs emerged within the four tissue-specific GRNs, and the GRN predictions matched expectations based upon published results for many of these examples. Our GRNs were also validated by ChIP-Seq datasets (KN1, FEA4 and O2). Key TFs were identified for each tissue and matched expectations for key regulators in each tissue, including GO enrichment and identity with known regulatory factors for that tissue. We also found functional modules in each network by clustering analysis with the MCL algorithm. CONCLUSIONS:By combining publicly available genome-wide expression data and network analysis, we can uncover GRNs at tissue-level resolution in maize. Since ChIP-Seq and PWMs are still limited in several model organisms, our study provides a uniform platform that can be adapted to any species with genome-wide expression data to construct GRNs. We also present a publicly available database, maize tissue-specific GRN (mGRN, https://www.bio.fsu.edu/mcginnislab/mgrn/ ), for easy querying. All source code and data are available at Github ( https://github.com/timedreamer/maize_tissue-specific_GRN ).

Project description:BackgroundPresent knowledge indicates a multilayered hierarchical gene regulatory network (ML-hGRN) often operates above a biological pathway. Although the ML-hGRN is very important for understanding how a pathway is regulated, there is almost no computational algorithm for directly constructing ML-hGRNs.ResultsA backward elimination random forest (BWERF) algorithm was developed for constructing the ML-hGRN operating above a biological pathway. For each pathway gene, the BWERF used a random forest model to calculate the importance values of all transcription factors (TFs) to this pathway gene recursively with a portion (e.g. 1/10) of least important TFs being excluded in each round of modeling, during which, the importance values of all TFs to the pathway gene were updated and ranked until only one TF was remained in the list. The above procedure, termed BWERF. After that, the importance values of a TF to all pathway genes were aggregated and fitted to a Gaussian mixture model to determine the TF retention for the regulatory layer immediately above the pathway layer. The acquired TFs at the secondary layer were then set to be the new bottom layer to infer the next upper layer, and this process was repeated until a ML-hGRN with the expected layers was obtained.ConclusionsBWERF improved the accuracy for constructing ML-hGRNs because it used backward elimination to exclude the noise genes, and aggregated the individual importance values for determining the TFs retention. We validated the BWERF by using it for constructing ML-hGRNs operating above mouse pluripotency maintenance pathway and Arabidopsis lignocellulosic pathway. Compared to GENIE3, BWERF showed an improvement in recognizing authentic TFs regulating a pathway. Compared to the bottom-up Gaussian graphical model algorithm we developed for constructing ML-hGRNs, the BWERF can construct ML-hGRNs with significantly reduced edges that enable biologists to choose the implicit edges for experimental validation.

Project description:BackgroundTranscription factors (TFs) and miRNAs are essential for the regulation of gene expression; however, the global view of human gene regulatory networks remains poorly understood. For example, how is the expression of so many genes regulated by limited cohorts of regulators and how are genes differentially expressed in different tissues despite the genetic code being the same in all tissues?ResultsWe analyzed the network properties of housekeeping and tissue-specific genes in gene regulatory networks from seven human tissues. Our results show that different classes of genes behave quite differently in these networks. Tissue-specific miRNAs show a higher average target number compared with non-tissue specific miRNAs, which indicates that tissue-specific miRNAs tend to regulate different sets of targets. Tissue-specific TFs exhibit higher in-degree, out-degree, cluster coefficient and betweenness values, indicating that they occupy central positions in the regulatory network and that they transfer genetic information from upstream genes to downstream genes more quickly than other TFs. Housekeeping TFs tend to have higher cluster coefficients compared with other genes that are neither housekeeping nor tissue specific, indicating that housekeeping TFs tend to regulate their targets synergistically. Several topological properties of disease-associated miRNAs and genes were found to be significantly different from those of non-disease-associated miRNAs and genes.ConclusionsTissue-specific miRNAs, TFs and disease genes have particular topological properties within the transcriptional regulatory networks of the seven human tissues examined. The tendency of tissue-specific miRNAs to regulate different sets of genes shows that a particular tissue-specific miRNA and its target gene set may form a regulatory module to execute particular functions in the process of tissue differentiation. The regulatory patterns of tissue-specific TFs reflect their vital role in regulatory networks and their importance to biological functions in their respective tissues. The topological differences between disease and non-disease genes may aid the discovery of new disease genes or drug targets. Determining the network properties of these regulatory factors will help define the basic principles of human gene regulation and the molecular mechanisms of disease.

Project description:Pilocytic Astrocytomas (PAs) are common low-grade central nervous system malignancies for which few recurrent and specific genetic alterations have been identified. In an effort to better understand the molecular biology underlying the pathogenesis of these pediatric brain tumors, we performed higher-order transcriptional network analysis of a large gene expression dataset to identify gene regulatory pathways that are specific to this tumor type, relative to other, more aggressive glial or histologically distinct brain tumours.RNA derived from frozen human PA tumours was subjected to microarray-based gene expression profiling, using Affymetrix U133Plus2 GeneChip microarrays. This data set was compared to similar data sets previously generated from non-malignant human brain tissue and other brain tumour types, after appropriate normalization.In this study, we examined gene expression in 66 PA tumors compared to 15 non-malignant cortical brain tissues, and identified 792 genes that demonstrated consistent differential expression between independent sets of PA and non-malignant specimens. From this entire 792 gene set, we used the previously described PAP tool to assemble a core transcriptional regulatory network composed of 6 transcription factor genes (TFs) and 24 target genes, for a total of 55 interactions. A similar analysis of oligodendroglioma and glioblastoma multiforme (GBM) gene expression data sets identified distinct, but overlapping, networks. Most importantly, comparison of each of the brain tumor type-specific networks revealed a network unique to PA that included repressed expression of ONECUT2, a gene frequently methylated in other tumor types, and 13 other uniquely predicted TF-gene interactions.These results suggest specific transcriptional pathways that may operate to create the unique molecular phenotype of PA and thus opportunities for corresponding targeted therapeutic intervention. Moreover, this study also demonstrates how integration of gene expression data with TF-gene and TF-TF interaction data is a powerful approach to generating testable hypotheses to better understand cell-type specific genetic programs relevant to cancer.

Dataset Information

Constructing tissue-specific transcriptional regulatory networks via a Markov random field.

Publications

Constructing tissue-specific transcriptional regulatory networks via a Markov random field.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets