Tissue-specific Protein-protein Interaction Networks Research Articles

A machine learning method for the identification and characterization of novel COVID-19 drug targets

In addition to vaccines, the World Health Organization sees novel medications as an urgent matter to fight the ongoing COVID-19 pandemic. One possible strategy is to identify target proteins, for which a perturbation by an existing compound is likely to benefit COVID-19 patients. In order to contribute to this effort, we present GuiltyTargets-COVID-19 (https://guiltytargets-covid.eu/), a machine learning supported web tool to identify novel candidate drug targets. Using six bulk and three single cell RNA-Seq datasets, together with a lung tissue specific protein-protein interaction network, we demonstrate that GuiltyTargets-COVID-19 is capable of (i) prioritizing meaningful target candidates and assessing their druggability, (ii) unraveling their linkage to known disease mechanisms, (iii) mapping ligands from the ChEMBL database to the identified targets, and (iv) pointing out potential side effects in the case that the mapped ligands correspond to approved drugs. Our example analyses identified 4 potential drug targets from the datasets: AKT3 from both the bulk and single cell RNA-Seq data as well as AKT2, MLKL, and MAPK11 in the single cell experiments. Altogether, we believe that our web tool will facilitate future target identification and drug development for COVID-19, notably in a cell type and tissue specific manner.

Detecting Rewiring Events in Protein-Protein Interaction Networks Based on Transcriptomic Data.

Proteins rarely carry out their cellular functions in isolation. Instead, eukaryotic proteins engage in about six interactions with other proteins on average. The aggregated protein interactome of an organism forms a “hairy ball”-type protein-protein interaction (PPI) network. Yet, in a typical human cell, only about half of all proteins are expressed at a particular time. Hence, it has become common practice to prune the full PPI network to the subset of expressed proteins. If RNAseq data is available, one can further resolve the specific protein isoforms present in a cell or tissue. Here, we review various approaches, software tools and webservices that enable users to construct context-specific or tissue-specific PPI networks and how these are rewired between two cellular conditions. We illustrate their different functionalities on the example of the interactions involving the human TNR6 protein. In an outlook, we describe how PPI networks may be integrated with epigenetic data or with data on the activity of splicing factors.

Compound Target Identification in Tissue-Specific Interaction Networks

Characterizing all possible side effects of compounds is not a trivial task due to unknown off-target proteins that might eventually lead lethal reactions. There is still a tremendous need of computational methods to identify protein targets of a new compound. We have performed a comprehensive analysis for identification of protein targets by integrating tissue-specific protein-protein interaction (PPI) networks and compound induced transcriptome data. Several network centrality metrics are computed to suggest the most probable off-targets of a given compound. The effects of interaction types between proteins and tissue-specific PPI networks are evaluated from multiple perspectives. Usage of network centrality metrics on a tissue-specific PPI network enhances the correct prediction rate of known targets of a given compound. The detailed analysis of successfully identified known targets indicated that degree and local radiality metrics are more practical for determining different types of target protein families, such as GPCR, chemokines, proteasome, and protein kinase families. Therefore, the proposed computational pipeline is applicable while investigating proteins from these families that can be especially targeted for treatment of complex diseases. To the best of our knowledge, this study for the first time presents how tissue-specific transcriptome changes alter the topological structure of PPI networks and the relative effects of tissue-specific networks in target identification. Overall, the proposed computational methods are practical tools for choosing more accurate protein targets for later computational analysis and wet-lab experiments.

NetworkAnalyst 3.0: a visual analytics platform for comprehensive gene expression profiling and meta-analysis.

The growing application of gene expression profiling demands powerful yet user-friendly bioinformatics tools to support systems-level data understanding. NetworkAnalyst was first released in 2014 to address the key need for interpreting gene expression data within the context of protein-protein interaction (PPI) networks. It was soon updated for gene expression meta-analysis with improved workflow and performance. Over the years, NetworkAnalyst has been continuously updated based on community feedback and technology progresses. Users can now perform gene expression profiling for 17 different species. In addition to generic PPI networks, users can now create cell-type or tissue specific PPI networks, gene regulatory networks, gene co-expression networks as well as networks for toxicogenomics and pharmacogenomics studies. The resulting networks can be customized and explored in 2D, 3D as well as Virtual Reality (VR) space. For meta-analysis, users can now visually compare multiple gene lists through interactive heatmaps, enrichment networks, Venn diagrams or chord diagrams. In addition, users have the option to create their own data analysis projects, which can be saved and resumed at a later time. These new features are released together as NetworkAnalyst 3.0, freely available at https://www.networkanalyst.ca.

Tissue-Specific Subcellular Localization Prediction Using Multi-Label Markov Random Fields.

The understanding of subcellular localization (SCL) of proteins and proteome variation in the different tissues and organs of the human body are two crucial aspects for increasing our knowledge of the dynamic rules of proteins, the cell biology, and the mechanism of diseases. Although there have been tremendous contributions to these two fields independently, the lack of knowledge of the variation of spatial distribution of proteins in the different tissues still exists. Here, we proposed an approach that allows predicting protein SCL on tissue specificity through the use of tissue-specific functional associations and physical protein-protein interactions (PPIs). We applied our previously developed Bayesian collective Markov random fields (BCMRFs) on tissue-specific protein-protein interaction network (PPI network) for nine types of tissues focusing on eight high-level SCL. The evaluated results demonstrate the strength of our approach in predicting tissue-specific SCL. We identified 1,314 proteins that their SCL were previously proven cell line dependent. We predicted 549 novel tissue-specific localized candidate proteins while some of them were validated via text-mining.

INBIA: a boosting methodology for proteomic network inference

BackgroundThe analysis of tissue-specific protein interaction networks and their functional enrichment in pathological and normal tissues provides insights on the etiology of diseases. The Pan-cancer proteomic project, in The Cancer Genome Atlas, collects protein expressions in human cancers and it is a reference resource for the functional study of cancers. However, established protocols to infer interaction networks from protein expressions are still missing.ResultsWe have developed a methodology called Inference Network Based on iRefIndex Analysis (INBIA) to accurately correlate proteomic inferred relations to protein-protein interaction (PPI) networks. INBIA makes use of 14 network inference methods on protein expressions related to 16 cancer types. It uses as reference model the iRefIndex human PPI network. Predictions are validated through non-interacting and tissue specific PPI networks resources. The first, Negatome, takes into account likely non-interacting proteins by combining both structure properties and literature mining. The latter, TissueNet and GIANT, report experimentally verified PPIs in more than 50 human tissues. The reliability of the proposed methodology is assessed by comparing INBIA with PERA, a tool which infers protein interaction networks from Pathway Commons, by both functional and topological analysis.ConclusionResults show that INBIA is a valuable approach to predict proteomic interactions in pathological conditions starting from the current knowledge of human protein interactions.

Integrating phenotypic features and tissue-specific information to prioritize disease genes

Prioritization of candidate disease genes is crucial for improving medical care, and is one of the fundamental challenges in the post-genomic era. In recent years, different network-based methods for gene prioritization are proposed. Previous studies on gene prioritization show that tissue-specific protein-protein interaction (PPI) networks built by integrating PPIs with tissue-specific gene expression profiles can perform better than tissue-na¨ive global PPI network. Based on the observations that diseases with similar phenotypes are likely to have common related genes, and genes associated with the same phenotype tend to interact with each other, we propose a method to prioritize disease genes based on a heterogeneous network built by integrating phenotypic features and tissue-specific information. In this heterogeneous network, the PPI network is built by integrating phenotypic features with a tissue-specific PPI network, and the disease network consists of the diseases that are associated with the same phenotype and tissue as the query disease. To determine the impacts of these two factors on gene prioritization, we test three typical network-based prioritization methods on heterogeneous networks consisting of combinations of different PPIs and disease networks built with or without phenotypic features and tissue-specific information. We also compare the proposed method with other tissuespecific networks. The results of case studies reveals that integrating phenotypic features with a tissue-specific PPI network improves the prioritization results. Moreover, the disease networks generated using our method not only show comparable performance with the widely used disease similarity dataset of 5080 human diseases, but are also effective for diseases that are not in the dataset.

SPECTRA: An Integrated Knowledge Base for Comparing Tissue and Tumor-Specific PPI Networks in Human.

Protein–protein interaction (PPI) networks available in public repositories usually represent relationships between proteins within the cell. They ignore the specific set of tissues or tumors where the interactions take place. Indeed, proteins can form tissue-selective complexes, while they remain inactive in other tissues. For these reasons, a great attention has been recently paid to tissue-specific PPI networks, in which nodes are proteins of the global PPI network whose corresponding genes are preferentially expressed in specific tissues. In this paper, we present SPECTRA, a knowledge base to build and compare tissue or tumor-specific PPI networks. SPECTRA integrates gene expression and protein interaction data from the most authoritative online repositories. We also provide tools for visualizing and comparing such networks, in order to identify the expression and interaction changes of proteins across tissues, or between the normal and pathological states of the same tissue. SPECTRA is available as a web server at http://alpha.dmi.unict.it/spectra.

Enhancing the Prioritization of Disease-Causing Genes through Tissue Specific Protein Interaction Networks

The prioritization of candidate disease-causing genes is a fundamental challenge in the post-genomic era. Current state of the art methods exploit a protein-protein interaction (PPI) network for this task. They are based on the observation that genes causing phenotypically-similar diseases tend to lie close to one another in a PPI network. However, to date, these methods have used a static picture of human PPIs, while diseases impact specific tissues in which the PPI networks may be dramatically different. Here, for the first time, we perform a large-scale assessment of the contribution of tissue-specific information to gene prioritization. By integrating tissue-specific gene expression data with PPI information, we construct tissue-specific PPI networks for 60 tissues and investigate their prioritization power. We find that tissue-specific PPI networks considerably improve the prioritization results compared to those obtained using a generic PPI network. Furthermore, they allow predicting novel disease-tissue associations, pointing to sub-clinical tissue effects that may escape early detection.