3plex enables deep computational investigation of triplex forming lncRNAs.

Long non-coding RNAs (lncRNAs) regulate gene expression through different molecular mechanisms, including DNA binding via the formation of RNA:DNA:DNA triple helices (TPXs). Despite the increasing amount of experimental evidence, TPXs investigation remains challenging. Here we present 3plex, a software able to predict TPX interactions in silico. Given an RNA sequence and a set of DNA sequences, 3plex integrates 1) Hoogsteen pairing rules that describe the biochemical interactions between RNA and DNA nucleotides, 2) RNA secondary structure prediction and 3) determination of the TPX thermal stability derived from a collection of TPX experimental evidences. We systematically collected and uniformly re-analysed published experimental lncRNA binding sites on human and mouse genomes. We used these data to evaluate 3plex performance and showed that its specific features allow a reliable identification of TPX interactions. We compared 3plex with the other available software and obtained comparable or even better accuracy at a fraction of the computation time. Interestingly, by inspecting collected data with 3plex we found that TPXs tend to be shorter and more degenerated than previously expected and that the majority of analysed lncRNAs can directly bind to the genome by TPX formation. Those results suggest that an important fraction of lncRNAs can exert its biological function through this mechanism. The software is available at https://github.com/molinerisLab/3plex.

Tissue memory relies on stem cell priming in distal undamaged areas.

Epithelial cells that participated in wound repair elicit a more efficient response to future injuries, which is believed to be locally restricted. Here we show that cell adaptation resulting from a localized tissue damage has a wide spatial impact at a scale not previously appreciated. We demonstrate that a specific stem cell population, distant from the original injury, originates long-lasting wound memory progenitors residing in their own niche. Notably, these distal memory cells have not taken part in the first healing but become intrinsically pre-activated through priming. This cell state, maintained at the chromatin and transcriptional level, leads to an enhanced wound repair that is partially recapitulated through epigenetic perturbation. Importantly wound memory has long-term harmful consequences, exacerbating tumourigenesis. Overall, we show that sub-organ-scale adaptation to injury relies on spatially organized memory-dedicated progenitors, characterized by an actionable cell state that establishes an epigenetic field cancerization and predisposes to tumour onset.

SNAI1 is upregulated during muscle regeneration and represses FGF21 and ATF3 expression by directly binding their promoters.

During skeletal myogenesis, the zinc-finger transcription factors SNAI1 and SNAI2, are expressed in proliferating myoblasts and regulate the transition to terminally differentiated myotubes while repressing pro-differentiation genes. Here, we demonstrate that SNAI1 is upregulated in vivo during the early phase of muscle regeneration induced by bupivacaine injury. Using shRNA-mediated gene silencing in C2C12 myoblasts and whole-transcriptome microarray analysis, we identified a collection of genes belonging to the endoplasmic reticulum (ER) stress pathway whose expression, induced by myogenic differentiation, was upregulated in absence of SNAI1. Among these, key ER stress genes, such as Atf3, Ddit3/Chop, Hspa5/Bip, and Fgf21, a myokine involved in muscle differentiation, were strongly upregulated. Furthermore, by promoter mutant analysis and Chromatin immune precipitation assay, we demonstrated that SNAI1 represses Fgf21 and Atf3 in proliferating myoblasts by directly binding to multiple E boxes in their respective promoter regions. Together, these data describe a new regulatory mechanism of myogenic differentiation involving the direct repressive action of SNAI1 on ER stress and Fgf21 expression, ultimately contributing to maintaining the proliferative and undifferentiated state of myoblasts.

The acetyltransferase p300 is recruited in trans to multiple enhancer sites by lncSmad7.

The histone acetyltransferase p300 (also known as KAT3B) is a general transcriptional coactivator that introduces the H3K27ac mark on enhancers triggering their activation and gene transcription. Genome-wide screenings demonstrated that a large fraction of long non-coding RNAs (lncRNAs) plays a role in cellular processes and organ development although the underlying molecular mechanisms remain largely unclear (1,2). We found 122 lncRNAs that interacts directly with p300. In depth analysis of one of these, lncSmad7, is required to maintain ESC self-renewal and it interacts to the C-terminal domain of p300. lncSmad7 also contains predicted RNA-DNA Hoogsteen forming base pairing. Combined Chromatin Isolation by RNA precipitation followed by sequencing (ChIRP-seq) together with CRISPR/Cas9 mutagenesis of the target sites demonstrate that lncSmad7 binds and recruits p300 to enhancers in trans, to trigger enhancer acetylation and transcriptional activation of its target genes. Thus, these results unveil a new mechanism by which p300 is recruited to the genome.

Deep learning for genomics using Janggu.

In recent years, numerous applications have demonstrated the potential of deep learning for an improved understanding of biological processes. However, most deep learning tools developed so far are designed to address a specific question on a fixed dataset and/or by a fixed model architecture. Here we present Janggu, a python library facilitates deep learning for genomics applications, aiming to ease data acquisition and model evaluation. Among its key features are special dataset objects, which form a unified and flexible data acquisition and pre-processing framework for genomics data that enables streamlining of future research applications through reusable components. Through a numpy-like interface, these dataset objects are directly compatible with popular deep learning libraries, including keras or pytorch. Janggu offers the possibility to visualize predictions as genomic tracks or by exporting them to the bigWig format as well as utilities for keras-based models. We illustrate the functionality of Janggu on several deep learning genomics applications. First, we evaluate different model topologies for the task of predicting binding sites for the transcription factor JunD. Second, we demonstrate the framework on published models for predicting chromatin effects. Third, we show that promoter usage measured by CAGE can be predicted using DNase hypersensitivity, histone modifications and DNA sequence features. We improve the performance of these models due to a novel feature in Janggu that allows us to include high-order sequence features. We believe that Janggu will help to significantly reduce repetitive programming overhead for deep learning applications in genomics, and will enable computational biologists to rapidly assess biological hypotheses.

Kinome-wide identification of phosphorylation networks in eukaryotic proteomes.

Motivation: Signaling and metabolic pathways are finely regulated by a network of protein phosphorylation events. Unraveling the nature of this intricate network, composed of kinases, target proteins and their interactions, is therefore of crucial importance. Although thousands of kinase-specific phosphorylations (KsP) have been annotated in model organisms their kinase-target network is far from being complete, with less studied organisms lagging behind. Results: In this work, we achieved an automated and accurate identification of kinase domains, inferring the residues that most likely contribute to peptide specificity. We integrated this information with the target peptides of known human KsP to predict kinase-specific interactions in other eukaryotes through a deep neural network, outperforming similar methods. We analyzed the differential conservation of kinase specificity among eukaryotes revealing the high conservation of the specificity of tyrosine kinases. With this approach we discovered 1590 novel KsP of potential clinical relevance in the human proteome. Availability and implementation: http://akid.bio.uniroma2.it. Supplementary information: Supplementary data are available at Bioinformatics online.

Viral bioterrorism: Learning the lesson of Ebola virus in West Africa 2013-2015.

Among the potential biological agents suitable as a weapon, Ebola virus represents a major concern. Classified by the CDC as a category A biological agent, Ebola virus causes severe hemorrhagic fever, characterized by high case-fatality rate; to date, no vaccine or approved therapy is available. The EVD epidemic, which broke out in West Africa since the late 2013, has got the issue of the possible use of Ebola virus as biological warfare agent (BWA) to come to the fore once again. In fact, due to its high case-fatality rate, population currently associates this pathogen to a real and tangible threat. Therefore, its use as biological agent by terrorist groups with offensive purpose could have serious repercussions from a psychosocial point of view as well as on closely sanitary level. In this paper, after an initial study of the main characteristics of Ebola virus, its potential as a BWA was evaluated. Furthermore, given the spread of the epidemic in West Africa in 2014 and 2015, the potential dissemination of the virus from an urban setting was evaluated. Finally, it was considered the actual possibility to use this agent as BWA in different scenarios, and the potential effects on one or more nation's stability.

Ebola virus disease 2013-2014 outbreak in west Africa: an analysis of the epidemic spread and response.

The Ebola virus epidemic burst in West Africa in late 2013, started in Guinea, reached in a few months an alarming diffusion, actually involving several countries (Liberia, Sierra Leone, Nigeria, Senegal, and Mali). Guinea and Liberia, the first nations affected by the outbreak, have put in place measures to contain the spread, supported by international organizations; then they were followed by the other nations affected. In the present EVD outbreak, the geographical spread of the virus has followed a new route: the achievement of large urban areas at an early stage of the epidemic has led to an unprecedented diffusion, featuring the largest outbreak of EVD of all time. This has caused significant concerns all over the world: the potential reaching of far countries from endemic areas, mainly through fast transports, induced several countries to issue information documents and health supervision for individuals going to or coming from the areas at risk. In this paper the geographical spread of the epidemic was analyzed, assessing the sequential appearance of cases by geographic area, considering the increase in cases and mortality according to affected nations. The measures implemented by each government and international organizations to contain the outbreak, and their effectiveness, were also evaluated.