Guillaume Devailly

CpG island erosion, polycomb occupancy and sequence motif enrichment at bivalent promoters in mammalian embryonic stem cells.

In embryonic stem (ES) cells, developmental regulators have a characteristic bivalent chromatin signature marked by simultaneous presence of both activation (H3K4me3) and repression (H3K27me3) signals and are thought to be in a ‘poised’ state for subsequent activation or silencing during differentiation. We collected eleven pairs (H3K4me3 and H3K27me3) of ChIP sequencing datasets in human ES cells and eight pairs in murine ES cells, and predicted high-confidence (HC) bivalent promoters. Over 85% of H3K27me3 marked promoters were bivalent in human and mouse ES cells. We found that (i) HC bivalent promoters were enriched for developmental factors and were highly likely to be differentially expressed upon transcription factor perturbation; (ii) murine HC bivalent promoters were occupied by both polycomb repressive component classes (PRC1 and PRC2) and grouped into four distinct clusters with different biological functions; (iii) HC bivalent and active promoters were CpG rich while H3K27me3-only promoters lacked CpG islands. Binding enrichment of distinct sets of regulators distinguished bivalent from active promoters. Moreover, a ‘TCCCC’ sequence motif was specifically enriched in bivalent promoters. Finally, this analysis will serve as a resource for future studies to further understand transcriptional regulation during embryonic development.

Gene expression variability in mammalian embryonic stem cells using single cell RNA-seq data

Background Gene expression heterogeneity contributes to development as well as disease progression. Due to technological limitations, most studies to date have focused on differences in mean expression across experimental conditions, rather than differences in gene expression variance. The advent of single cell RNA sequencing has now made it feasible to study gene expression heterogeneity and to characterise genes based on their coefficient of variation. Methods We collected single cell gene expression profiles for 32 human and 39 mouse embryonic stem cells and studied correlation between diverse characteristics such as network connectivity and coefficient of variation (CV) across single cells. We further systematically characterised properties unique to High CV genes. Results Highly expressed genes tended to have a low CV and were enriched for cell cycle genes. In contrast, High CV genes were co-expressed with other High CV genes, were enriched for bivalent (H3K4me3 and H3K27me3) marked promoters and showed enrichment for response to DNA damage and DNA repair. Conclusions Taken together, this analysis demonstrates the divergent characteristics of genes based on their CV. High CV genes tend to form co-expression clusters and they explain bivalency at least in part.

Variable reproducibility in genome-scale public data: A case study using ENCODE ChIP sequencing resource.

Genome‐wide data is accumulating in an unprecedented way in the public domain. Re‐mining this data shows great potential to generate novel hypotheses. However this approach is dependent on the quality (technical and biological) of the underlying data. Here we performed a systematic analysis of chromatin immunoprecipitation (ChIP) sequencing data of transcription and epigenetic factors from the encyclopaedia of DNA elements (ENCODE) resource to demonstrate that about one third of conditions with replicates show low concordance between replicate peak lists. This serves as a case study to demonstrate a caveat concerning genome‐wide analyses and highlights a need to validate the quality of each sample before performing further associative analyses.

The PLA2R1-JAK2 pathway upregulates ERRα and its mitochondrial program to exert tumor-suppressive action.

Little is known about the biological role of the phospholipase A2 receptor (PLA2R1) transmembrane protein. In recent years, PLA2R1 has been shown to have an important role in regulating tumor-suppressive responses via JAK2 activation, but the underlying mechanisms are largely undeciphered. In this study, we observed that PLA2R1 increases the mitochondrial content, judged by increased levels of numerous mitochondrial proteins, of the mitochondrial structural component cardiolipin, of the mitochondrial DNA content, and of the mitochondrial DNA replication and transcription factor TFAM. This effect of PLA2R1 relies on a transcriptional program controlled by the estrogen-related receptor alpha1 (ERRα) mitochondrial master regulator. Expression of ERRα and of its nucleus-encoded mitochondrial targets is upregulated upon PLA2R1 ectopic expression, and this effect is mediated by JAK2. Conversely, downregulation of PLA2R1 decreases the level of ERRα and of its nucleus-encoded mitochondrial targets. Finally, blocking the ERRα-controlled mitochondrial program largely inhibits the PLA2R1-induced tumor-suppressive response. Together, our data document ERRα and its mitochondrial program as downstream effectors of the PLA2R1-JAK2 pathway leading to oncosuppression.

Heat*seq: an interactive web tool for high-throughput sequencing experiment comparison with public data

Summary: Better protocols and decreasing costs have made high-throughput sequencing experiments now accessible even to small experimental laboratories. However, comparing one or few experiments generated by an individual lab to the vast amount of relevant data freely available in the public domain might be limited due to lack of bioinformatics expertise. Though several tools, including genome browsers, allow such comparison at a single gene level, they do not provide a genome-wide view. We developed Heat*seq, a web-tool that allows genome scale comparison of high throughput experiments chromatin immuno-precipitation followed by sequencing, RNA-sequencing and Cap Analysis of Gene Expression) provided by a user, to the data in the public domain. Heat*seq currently contains over 12 000 experiments across diverse tissues and cell types in human, mouse and drosophila. Heat*seq displays interactive correlation heatmaps, with an ability to dynamically subset datasets to contextualize user experiments. High quality figures and tables are produced and can be downloaded in multiple formats. Availability and Implementation: Web application: http://www.heatstarseq.roslin.ed.ac.uk/. Source code: https://github.com/gdevailly. Contact: Guillaume.Devailly@roslin.ed.ac.uk or Anagha.Joshi@roslin.ed.ac.uk Supplementary information: Supplementary data are available at Bioinformatics online.

Delineating biological and technical variance in single cell expression data

Single cell transcriptomics is becoming a common technique to unravel new biological phenomena whose functional significance can only be understood in the light of differences in gene expression between single cells. The technology is still in its early days and therefore suffers from many technical challenges. This review discusses the continuous effort to identify and systematically characterise various sources of technical variability in single cell expression data and the need to further develop experimental and computational tools and resources to help deal with it.

scFeatureFilter: Correlation-Based Feature Filtering for Single-Cell RNAseq

Single cell RNA sequencing is becoming increasingly popular due to rapidly evolving technology, decreasing costs and its wide applicability. However, the technology suffers from high drop-out rate and high technical noise, mainly due to the low starting material. This hinders the extraction of biological variability, or signal, from the data. One of the first steps in the single cell analysis pipelines is, therefore, to filter the data to keep the most informative features only. This filtering step is often done by arbitrarily selecting a threshold.

Dynamics of promoter bivalency and RNAP II pausing in mouse stem and differentiated cells

BackgroundMammalian embryonic stem cells display a unique epigenetic and transcriptional state to facilitate pluripotency by maintaining lineage-specification genes in a poised state. Two epigenetic and transcription processes involved in maintaining poised state are bivalent chromatin, characterized by the simultaneous presence of activating and repressive histone methylation marks, and RNA polymerase II (RNAPII) promoter proximal pausing. However, the dynamics of histone modifications and RNAPII at promoters in diverse cellular contexts remains underexplored.ResultsWe collected genome wide data for bivalent chromatin marks H3K4me3 and H3K27me3, and RNAPII (8WG16) occupancy together with expression profiling in eight different cell types, including ESCs, in mouse. The epigenetic and transcription profiles at promoters grouped in over thirty clusters with distinct functional identities and transcription control.ConclusionThe clustering analysis identified distinct bivalent clusters where genes in one cluster retained bivalency across cell types while in the other were mostly cell type specific, but neither showed a high RNAPII pausing. We noted that RNAPII pausing is more associated with active genes than bivalent genes in a cell type, and was globally reduced in differentiated cell types compared to multipotent.

Transcription Control in Human Cell Types by Systematic Analysis of ChIP Sequencing Data from the ENCODE

Transcription control plays a key role during development and disease with trans-acting factors (TFs) regulating expression of genes through DNA interaction. ChIP sequencing is widely used to get the genome wide binding profiles of TFs in a cell type of interest. The reduction in cost of sequencing and the technological improvement has resulted in vast amount of ChIP sequencing data accumulating in the public domain. The ENCODE consortium alone provides 690 publicly available ChIP sequencing data sets across 91 human cell types. We performed a multi-facetted bioinformatics analysis of this data to unravel diverse properties of TFs in the cellular context. Specifically, we characterised genomic location as well as sequence motif preference of the factors. We demonstrated that the distal binding of factors is more cell type specific than the promoter proximal. We identified combinations of factors acting in concert at distinct genomic loci. Finally, we highlighted how this data is of value to associate novel regulators to disease by integrating it with disease-associated gene loci obtained from GWAS studies.

Meta-Analysis Of Liver Transcriptomic Data To Identify Mammalian Functional Orthologs

Functional annotation transfer across multi-gene family orthologs can lead to functional misannotations. We hypothesised that co-expression network will help predict functional orthologs amongst complex homologous gene families. To explore the use of transcriptomic data available in public domain to identify functionally equivalent ones from all predicted orthologs, we collected genome wide expression data in mouse and rat liver from over 1500 experiments with varied treatments. We used a hyper-graph clustering method to identify clusters of orthologous genes co-expressed in both mouse and rat. We validated these clusters by analysing expression profiles in each species separately, and demonstrating a high overlap. We then focused on genes in 18 homology groups with one-to-many or many-to-many relationships between two species, to discriminate between functionally equivalent and non-equivalent orthologs. Finally, we further applied our method by collecting heart transcriptomic data (over 1400 experiments) in rat and mouse to validate the method in an independent tissue.Identifying which orthologs share functions from sequence alone can be challenging, notably in case of paralogous genes families. We hypothesised that co-expression network will help predict functional orthologs amongst complex homologous gene families. To explore the use of transcriptomic data available in public domain to predict functionally equivalent orthologs, we collected genome wide expression data in mouse and rat liver from over 1500 experiments with varied treatments. We used a hyper-graph clustering method to identify clusters of orthologous genes co-expressed in both mouse and rat. We validated these clusters by analysing expression profiles in each species separately, and demonstrating a high overlap. We then focused on genes in 18 homology groups with one-to-many or many-to-many relationships between two species, to discriminate between functionally equivalent and non-equivalent orthologs.