Long RNA-Mediated Chromatin Regulation in Fission Yeast and Mammals

As part of a complex network of genome control, long regulatory RNAs exert significant influences on chromatin dynamics. Understanding how this occurs could illuminate new avenues for disease treatment and lead to new hypotheses that would advance gene regulatory research. Recent studies using the model fission yeast Schizosaccharomyces pombe (S. pombe) and powerful parallel sequencing technologies have provided many insights in this area. This review will give an overview of key findings in S. pombe that relate long RNAs to multiple levels of chromatin regulation: histone modifications, gene neighborhood regulation in cis and higher-order chromosomal ordering. Moreover, we discuss parallels recently found in mammals to help bridge the knowledge gap between the study systems.

Daisychain: Search and Interactive Visualisation of Homologs in Genome Assemblies

Daisychain is an interactive graph visualisation and search tool for custom-built gene homology databases. The main goal of Daisychain is to allow researchers working with specific genes to identify homologs in other annotation releases. The gene-centric representation includes local gene neighborhood to distinguish orthologs and paralogs by local synteny. The software supports genome sequences in FASTA format and GFF3 formatted annotation files, and the process of building the homology database requires a minimum amount of user interaction. Daisychain includes an integrated web viewer that can be used for both data analysis and data publishing. The web interface extends KnetMaps.js and is based on JavaScript.

Detecting operons in bacterial genomes via visual representation learning

Contiguous genes in prokaryotes are often arranged into operons. Detecting operons plays a critical role in inferring gene functionality and regulatory networks. Human experts annotate operons by visually inspecting gene neighborhoods across pileups of related genomes. These visual representations capture the inter-genic distance, strand direction, gene size, functional relatedness, and gene neighborhood conservation, which are the most prominent operon features mentioned in the literature. By studying these features, an expert can then decide whether a genomic region is part of an operon. We propose a deep learning based method named Operon Hunter that uses visual representations of genomic fragments to make operon predictions. Using transfer learning and data augmentation techniques facilitates leveraging the powerful neural networks trained on image datasets by re-training them on a more limited dataset of extensively validated operons. Our method outperforms the previously reported state-of-the-art tools, especially when it comes to predicting full operons and their boundaries accurately. Furthermore, our approach makes it possible to visually identify the features influencing the network’s decisions to be subsequently cross-checked by human experts.

TREND: a platform for exploring protein function in prokaryotes based on phylogenetic, domain architecture and gene neighborhood analyses

Key steps in a computational study of protein function involve analysis of (i) relationships between homologous proteins, (ii) protein domain architecture and (iii) gene neighborhoods the corresponding proteins are encoded in. Each of these steps requires a separate computational task and sets of tools. Currently in order to relate protein features and gene neighborhoods information to phylogeny, researchers need to prepare all the necessary data and combine them by hand, which is time-consuming and error-prone. Here, we present a new platform, TREND (tree-based exploration of neighborhoods and domains), which can perform all the necessary steps in automated fashion and put the derived information into phylogenomic context, thus making evolutionary based protein function analysis more efficient. A rich set of adjustable components allows a user to run the computational steps specific to his task. TREND is freely available at http://trend.zhulinlab.org.

Detecting Operons in Bacterial Genomes via Visual Representation Learning

ABSTRACTContiguous genes in prokaryotes are often arranged into operons. Detecting operons plays a critical role in inferring gene functionality and regulatory networks. Human experts annotate operons by visually inspecting gene neighborhoods across pileups of related genomes. These visual representations capture the inter-genic distance, strand direction, gene size, functional relatedness, and gene neighborhood conservation, which are the most prominent operon features mentioned in the literature. By studying these features, an expert can then decide whether a genomic region is part of an operon. We propose a deep learning based method named Operon Hunter that uses visual representations of genomic fragments to make operon predictions. Using transfer learning and data augmentation techniques facilitates leveraging the powerful neural networks trained on image datasets by re-training them on a more limited dataset of extensively validated operons. Our method outperforms the previously reported state-of-the-art tools, especially when it comes to predicting full operons and their boundaries accurately. Furthermore, our approach makes it possible to visually identify the features influencing the network’s decisions to be subsequently cross-checked by human experts.

Compound signaling activates endogenous retroviruses by inducing enhancer and gene-neighborhood transcription

SummaryMultiple sclerosis (MS) is a neuroinflammatory and autoimmune disease, in which various immune cell types and autoreactive T cells exert a pathogenic activity. This disease is also associated with increased transcription of several endogenous retroviruses (HERVs) normally kept in check by heterochromatin. Here, we have uncovered an organic pollutant dieldrin that activates several HERVs associated with MS and allowing us to examine the mechanism of their activation. Dieldrin singles out by its ability to simultaneously activate the MAP kinase and the PI3K pathways, while also triggering calcium dependent peptidylarginine deiminase activity. It was this association of pathways that caused HERV activation, a phenomenon that was only part of more generally increased transcription of heterochromatic regions. The HERV transcripts were generally not polyadenylated. Some arose as a consequence of activation of HERV-based enhancers, while others were the result of unusually strong activation at some mostly transcription factor genes causing transcription to leak out of the HERV-free region that surrounds them. Altogether, our data emphasized the hazard associated with simultaneous activation of multiples signaling pathways by xenobiotics, while also providing a very general toolbox for the interpretation of HERV transcription.