Biology subjects

Backofen, R.

Publications and source records attributed to Backofen, R..

53 records · Page 3Linked to original sources

bioRxiv2020

A global data-driven census of Salmonella small proteins and their potential functions in bacterial virulence

Small proteins are an emerging class of gene products with diverse roles in bacterial physiology. However, a full understanding of their importance has been hampered by insufficient genome annotations and a lack of comprehensive characterization in microbes other than Escherichia coli. We have taken an integrative approach to accelerate the discovery of small proteins and their putative virulence-associated functions in Salmonella Typhimurium. We merged the annotated small proteome of Salmonella with new small proteins predicted with in silico and experimental approaches. We then exploited existing and newly generated global datasets that provide information on small open reading frame expression during infection of epithelial cells (dual RNA-seq), contribution to bacterial fitness inside macrophages (TraDIS), and potential engagement in molecular interactions (Grad-seq). This integrative approach suggested a new role for the small protein MgrB beyond its known function in regulating PhoQ. We demonstrate a virulence and motility defect of a Salmonella {Delta}mgrB mutant and reveal an effect of MgrB in regulating the Salmonella transcriptome and proteome under infection-relevant conditions. Our study highlights the power of interpreting available "omics" datasets with a focus on small proteins, and may serve as a blueprint for a data integration-based survey of small proteins in diverse bacteria.

microbiology↗

bioRxiv2020

HRIBO - High-throughput analysis of bacterial ribosome profiling data

MotivationRibosome profiling (Ribo-seq) is a powerful approach based on ribosome-protected RNA fragments to explore the translatome of a cell, and is especially useful for the detection of small proteins (<=70 amino acids) that are recalcitrant to biochemical and in silico approaches. While pipelines are available to analyze Ribo-seq data, none are designed explicitly for the analysis of Ribo-seq data from prokaryotes, nor are they focused on the discovery of unannotated open reading frames (ORFs) in bacteria. ResultsWe present HRIBO (High-throughput annotation by Ribo-seq), a workflow to enable reproducible and high-throughput analysis of bacterial Ribo-seq data. The workflow performs all required pre-processing and quality control steps. Importantly, HRIBO outputs annotation-independent ORF predictions based on two complementary bacteria-focused tools, and integrates them with additional features. This facilitates the rapid discovery of novel ORFs and their prioritization for functional characterization. AvailabilityHRIBO is a free and open source project available under the GPL-3 license at: https://github.com/RickGelhausen/HRIBO

bioinformatics↗

bioRxiv2020

Loop detection using Hi-C data with HiCExplorer

Chromatin loops are an important factor in the structural organization of the genome. The detection of chromatin loops in Hi-C interaction matrices is a challenging and compute intensive task. The presented approach shows a chromatin loop detection algorithm which applies a strict candidate selection based on continuous negative binomial distributions and performs a Wilcoxon rank-sum test to detect enriched Hi-C interactions.

bioinformatics↗

bioRxiv2020

Approximate k-nearest neighbors graph for single-cell Hi-C dimensional reduction with MinHash

Single-cell Hi-C interaction matrices are high dimensional and very sparse. To cluster thousands of single-cell Hi-C interaction matrices they are flattened and compiled into one matrix. This matrix can, depending on the resolution, have a few millions or even billions of features and any computation with it is therefore memory demanding. A common approach to reduce the number of features is to compute a nearest neighbors graph. However, the exact euclidean distance computation is in O(n2) and therefore we present an implementation of an approximate nearest neighbors method based on local sensitive hashing running in O(n). The presented method is able to process a 10kb single-cell Hi-C data set with 2500 cells and needs 53 GB of memory while the exact k-nearest neighbors approach is not computable with 1 TB of memory.

bioinformatics↗

bioRxiv2019

GraphProt2: A novel deep learning-based method for predicting binding sites of RNA-binding proteins

CLIP-seq is the state-of-the-art technique to experimentally determine transcriptome-wide binding sites of RNA-binding proteins (RBPs). However, it relies on gene expression which can be highly variable between conditions, and thus cannot provide a complete picture of the RBP binding landscape. This creates a demand for computational methods to predict missing binding sites. Here we present GraphProt2, a computational RBP binding site prediction framework based on graph convolutional neural networks (GCNs). In contrast to current CNN methods, GraphProt2 offers native support for the encoding of base pair information as well as variable length input, providing increased flexibility and the prediction of nucleotide-wise RBP binding profiles. We demonstrate its superior performance compared to GraphProt and two CNN-based methods on single as well as combined CLIP-seq datasets. Conceived as an end-to-end method, GraphProt2 includes all necessary functionalities, from dataset generation over model training to the evaluation of binding preferences and binding site prediction. Various input types and features are supported, accompanied by comprehensive statistics and visualizations to inform the user about datatset characteristics and learned model properties. All this makes GraphProt2 the most versatile and complete RBP binding site prediction method available so far.

bioinformatics↗

bioRxiv2019

Tool recommender system in Galaxy using deep learning

Galaxy is a web-based and open-source scientific data-processing platform. Researchers compose pipelines in Galaxy to analyse scientific data. These pipelines, also known as workflows, can be complex and difficult to create from thousands of tools, especially for researchers new to Galaxy. To make creating workflows easier, faster and less error-prone, a predictive system is developed to recommend tools facilitating further analysis. A model is created to recommend tools by analysing workflows, composed by researchers on the European Galaxy server, using a deep learning approach. The higher-order dependencies in workflows, represented as directed acyclic graphs, are learned by training a gated recurrent units (GRU) neural network, a variant of a recurrent neural network (RNN). The weights of tools used in the neural network training are derived from their usage frequencies over a period of time. The hyper-parameters of the neural network are optimised using Bayesian optimisation. An accuracy of 97% in predicting tools is achieved by the model for precision@1, precision@2 and precision@3 metrics. It is accessed by a Galaxy API to recommend tools in real-time. Multiple user interface (UI) integrations on the server communicate with this API to apprise researchers of these recommended tools interactively. Contactkumara@informatik.uni-freiburg.de gruening@informatik.uni-freiburg.de backofen@informatik.uni-freiburg.de

bioinformatics↗

bioRxiv2019

CRISPRCasIdentifier: Machine learning for accurate identification and classification of CRISPR-Cas systems

CRISPR-Cas genes are extraordinarily diverse and evolve rapidly when compared to other prokaryotic genes. With the rapid increase in newly sequenced archaeal and bacterial genomes, manual identification of CRISPR-Cas systems is no longer viable. Thus, an automated approach is required for advancing our understanding of the evolution and diversity of these systems, and for finding new candidates for genome engineering in eukaryotic models. In this paper, we introduce a holistic strategy that combines regression and classification models for improving the quality of protein cascades, predicting their subtypes, detecting signature genes and extracting potential rules that reveal functional modules for CRISPR.

bioinformatics↗

bioRxiv2019

StoatyDive: Evaluation and Classification of Peak Profiles for Sequencing Data

The prediction of binding sites (peak calling) is a common task in the data analysis of methods such as crosslinking or chromatin immunoprecipitation in combination with high-throughput sequencing (CLIP-Seq, ChIP-Seq). The predicted binding sites are often further analyzed to predict sequence motifs or structure patterns as an example. However, the obtained peak set can vary in their profile shapes because of the used peakcaller method, different binding domains of the protein, protocol biases, or other factors. Thus, a tool is missing that evaluates and classifies the predicted peaks based on their shapes. We hereby present StoatyDive, a tool that can be used to filter for specific peak profile shapes of sequencing data such as CLIP and ChIP. StoatyDive therefore fine tunes downstream analysis steps such as structure or sequence motif predictions and acts as a quality control.\n\nWith StoatyDive we were able to classify distinct peak profile shapes from CLIP-seq data of the histone stem-loop-binding protein (SLBP). We show the potential of StoatyDive, as a quality control tool and as a filter to pick different shapes based on biological or methodical questions.\n\nStoatyDive is open source and freely available under GLP-3 at https://github.com/BackofenLab/StoatyDive and at bioconda https://anaconda.org/bioconda/stoatydive.

bioinformatics↗

bioRxiv2019

Empowering the annotation and discovery of structured RNAs with scalable and accessible integrative clustering

RNA plays essential regulatory roles in all known forms of life. Clustering RNA sequences with common sequence and structure is an essential step towards studying RNA function. With the advent of high-throughput sequencing techniques, experimental and genomic data are expanding to complement the predictive methods. However, the existing methods do not effectively utilize and cope with the immense amount of data becoming available. Here we present GraphClust2, a comprehensive approach for scalable clustering of RNAs based on sequence and structural similarities. GraphClust2 provides an integrative solution by incorporating diverse types of experimental and genomic data in an accessible fashion via the Galaxy framework. We demonstrate that the tasks of clustering and annotation of structured RNAs can be considerably improved, through a scalable methodology that also supports structure probing data. Based on this, we further introduce an off-the-shelf procedure to identify locally conserved structure candidates in long RNAs. In this way, we suggest the presence and the sparsity of phylogenetically conserved local structures in some long non-coding RNAs. Furthermore, we demonstrate the advantage of a scalable clustering for discovering structured motifs under inherent and experimental biases and uncover prominent targets of the double-stranded RNA binding protein Roquin-1 that are evolutionary conserved.

bioinformatics↗

bioRxiv2018

uORF-Tools - Workflow for the determination of translation-regulatory upstream open reading frames

Ribosome profiling (ribo-seq) provides a means to analyze active translation by determining ribosome occupancy in a transcriptome-wide manner. The vast majority of ribosome protected fragments resides within the protein-coding sequence of mRNAs. However, commonly reads are also found within the transcript leader sequence (TLS) (aka 5 untranslated region) preceding the main open reading frame (ORF), which indicates the translation of regulatory upstream ORFs (uORFs). Here, we present a workflow for the identification of functional uORFs, which contribute to the translational regulation of their associated main ORFs. The workflow is available as free and open Snakemake workflow. Furthermore, we provide a comprehensive human uORF annotation file, which can be used within the pipeline, thus reducing the runtime. (Availability: https://github.com/anibunny12/uORF-Tools)

bioinformatics↗

bioRxiv2018

Structure probing data enhances RNA-RNA interaction prediction

SummaryExperimental structure probing data has been shown to improve thermodynamics-based RNA secondary structure prediction. To this end, chemical reactivity information (as provided e.g. by SHAPE) is incorporated, which encodes whether or not individual nucleotides are involved in intra-molecular structure. Since inter-molecular RNA-RNA interactions are often confined to unpaired RNA regions, SHAPE data is even more promising to improve interaction prediction. Here we show how such experimental data can be incorporated seamlessly into accessibility-based RNA-RNA interaction prediction approaches, as implemented in IntaRNA. This is possible via the computation and use of unpaired probabilities that incorporate the structure probing information. We show that experimental SHAPE data can significantly improve RNA-RNA interaction prediction. We evaluate our approach by investigating interactions of a spliceosomal U1 snRNA transcript with its target splice sites. When SHAPE data is incorporated, known target sites are predicted with increased precision and specificity.\n\nAvailabilityhttps://github.com/BackofenLab/IntaRNA

bioinformatics↗

bioRxiv2018

MechRNA: prediction of lncRNA mechanisms from RNA-RNA and RNA-protein interactions

MotivationLong non-coding RNAs (lncRNAs) are defined as transcripts longer than 200 nucleotides that do not get translated into proteins. Often these transcripts are processed (spliced, capped, polyadenylated) and some are known to have important biological functions. However, most lncRNAs have unknown or poorly understood functions. Nevertheless, because of their potential role in cancer, lncRNAs are receiving a lot of attention, and the need for computational tools to predict their possible mechanisms of action is more than ever. Fundamentally, most of the known lncRNA mechanisms involve RNA-RNA and/or RNA-protein interactions. Through accurate predictions of each kind of interaction and integration of these predictions, it is possible to elucidate potential mechanisms for a given lncRNA.\n\nApproachHere we introduce MechRNA, a pipeline for corroborating RNA-RNA interaction prediction and protein binding prediction for identifying possible lncRNA mechanisms involving specific targets or on a transcriptome-wide scale. The first stage uses a version of IntaRNA2 with added functionality for efficient prediction of RNA-RNA interactions with very long input sequences, allowing for large-scale analysis of lncRNA interactions with little or no loss of optimality. The second stage integrates protein binding information pre-computed by GraphProt, for both the lncRNA and the target. The final stage involves inferring the most likely mechanism for each lncRNA/target pair. This is achieved by generating candidate mechanisms from the predicted interactions, the relative locations of these interactions and correlation data, followed by selection of the most likely mechanistic explanation using a combined p-value.\n\nResultsWe applied MechRNA on a number of recently identified cancer-related lncRNAs (PCAT1, PCAT29, ARLnc1) and also on two well-studied lncRNAs (PCA3 and 7SL). This led to the identification of hundreds of high confidence potential targets for each lncRNA and corresponding mechanisms. These predictions include the known competitive mechanism of 7SL with HuR for binding on the tumor suppressor TP53, as well as mechanisms expanding what is known about PCAT1 and ARLn1 and their targets BRCA2 and AR, respectively. For PCAT1-BRCA2, the mechanism involves competitive binding with HuR, which we confirmed using HuR immunoprecipitation assays.\n\nAvailabilityMechRNA is available for download at https://bitbucket.org/compbio/mechrna\n\nContactbackofen@informatik.uni-freiburg.de, cenksahi@indiana.edu\n\nSupplementary informationSupplementary data are available at Bioinformatics online.

bioinformatics↗

bioRxiv2018

Conserved accessory proteins encoded with archaeal and bacterial Type III CRISPR-Cas gene cassettes that may specifically modulate, complement or extend interference activity

A study was undertaken to identify conserved proteins that are encoded either within, or directly adjacent to, cas gene cassettes of Type III CRISPR-Cas interference modules. These Type III modules are especially versatile functionally and have been shown to target and degrade dsDNA, ssDNA and ssRNA. In addition, the interference gene cassettes are frequently intertwined with other accessory genes, including genes encoding CARF domains, some of which are likely to be cofunctional. Using a comparative genomics approach, and defining a Type III association score accounting for coevolution and specificity of flanking genes, we identified and classified 39 new Type III associated gene families. Most archaeal and bacterial Type III modules were seen to be flanked by several accessory genes, around half of which did not encode CARF domains and remain of unknown function. Non-CARF accessory genes were found to be more diverse than their CARF counterparts, encoding nuclease, helicase, protease, ATPase, transporter and transmembrane domains and including a considerable fraction that encoded no known domains. The diversity of non-CARF Type III accessory genes found in this study suggests that additional families exist which remain undetected because of the limited number of annotated genomes currently available. The method employed is scalable for potential application on metagenomic data once automated pipelines for annotation of CRISPR-Cas systems have been developed. All accessory genes found in this study are presented online in a readily accessible and searchable format for researchers to audit their model organism of choice: http://accessory.crispr.dk.

bioinformatics↗

bioRxiv2017

Community-driven data analysis training for biology

The primary problem with the explosion of biomedical datasets is not the data itself, not computational resources, and not the required storage space, but the general lack of trained and skilled researchers to manipulate and analyze these data. Eliminating this problem requires development of comprehensive educational resources. Here we present a community-driven framework that enables modern, interactive teaching of data analytics in life sciences and facilitates the development of training materials. The key feature of our system is that it is not a static but a continuously improved collection of tutorials. By coupling tutorials with a web-based analysis framework, biomedical researchers can learn by performing computation themselves through a web-browser without the need to install software or search for example datasets. Our ultimate goal is to expand the breadth of training materials to include fundamental statistical and data science topics and to precipitate a complete re-engineering of undergraduate and graduate curricula in life sciences.

bioinformatics↗

bioRxiv2017

Distinct epigenetic programs regulate cardiac myocyte development and disease in the human heart in vivo

Epigenetic mechanisms and transcription factor networks essential for differentiation of cardiac myocytes have been uncovered. However, reshaping of the epigenome of these terminally differentiated cells during fetal development, postnatal maturation and in disease remains unknown. Thus, the aim of this study was to determine the dynamics of the cardiac myocyte epigenome during development and in chronic heart failure. Prenatal development and postnatal maturation are characterized by a cooperation of active CpG methylation and histone marks at cis-regulatory and genic regions to shape the cardiac myocyte transcriptome. In contrast, pathological gene expression in terminal heart failure is accompanied by changes in active histone marks without major alterations in CpG methylation and repressive chromatin marks. Notably, cis-regulatory regions in cardiac myocytes are significantly enriched for cardiovascular disease-associated variants. This study uncovers distinct layers of epigenetic regulation not only during prenatal development and postnatal maturation but also in diseased human cardiac myocytes.

genomics↗

bioRxiv2017

Practical computational reproducibility in the life sciences

Many areas of research suffer from poor reproducibility. This problem is particularly acute in computationally intensive domains where results rely on a series of complex methodological decisions that are not well captured by traditional publication approaches. Various guidelines have emerged for achieving reproducibility, but practical implementation of these practices remains difficult. This is because reproducing published computational analyses requires installing many software tools plus associated libraries, connecting tools together into the complete pipeline, and specifying parameters. Here we present a suite of recently emerged technologies which make computational reproducibility not just possible, but, finally, practical in both time and effort. By combining a system for building highly portable packages of bioinformatics software, containerization and virtualization technologies for isolating reusable execution environments for these packages, and an integrated workflow system that automatically orchestrates the composition of these packages for entire pipelines, an unprecedented level of computational reproducibility can be achieved.

bioinformatics↗

bioRxiv2017

uvCLAP: a fast, non-radioactive method to identify in vivo targets of RNA-binding proteins

RNA-binding proteins (RBPs) play important and essential roles in eukaryotic gene expression regulating splicing, localization, translation and stability of mRNAs. Understanding the exact contribution of RBPs to gene regulation is crucial as many RBPs are frequently mis-regulated in several neurological diseases and certain cancers. While recently developed techniques provide binding sites of RBPs, they are labor-intensive and generally rely on radioactive labeling of RNA. With more than 1,000 RBPs in a human cell, it is imperative to develop easy, robust, reproducible and high-throughput methods to determine in vivo targets of RBPs. To address these issues we developed uvCLAP (UV crosslinking and affinity purification) as a robust, reproducible method to measure RNA-protein interactions in vivo. To test its performance and applicability we investigated binding of 15 RBPs from fly, mouse and human cells. We show that uvCLAP generates reliable and comparable data to other methods. Unexpectedly, our results show that despite their different subcellular localizations, STAR proteins (KHDRBS1-3, QKI) bind to a similar RNA motif in vivo. Consistently a point mutation (KHDRBS1Y440F) or a natural splice isoform (QKI-6) that changes the respective RBP subcellular localization, dramatically alters target selection without changing the targeted RNA motif. Combined with the knowledge that RBPs can compete and cooperate for binding sites, our data shows that compartmentalization of RBPs can be used as an elegant means to generate RNA target specificity.

molecular biology↗