Reconstruction Of Protein-protein Interaction Networks Research Articles

From complete cross-docking to partners identification and binding sites predictions.

Proteins ensure their biological functions by interacting with each other. Hence, characterising protein interactions is fundamental for our understanding of the cellular machinery, and for improving medicine and bioengineering. Over the past years, a large body of experimental data has been accumulated on who interacts with whom and in what manner. However, these data are highly heterogeneous and sometimes contradictory, noisy, and biased. Ab initio methods provide a means to a "blind" protein-protein interaction network reconstruction. Here, we report on a molecular cross-docking-based approach for the identification of protein partners. The docking algorithm uses a coarse-grained representation of the protein structures and treats them as rigid bodies. We applied the approach to a few hundred of proteins, in the unbound conformations, and we systematically investigated the influence of several key ingredients, such as the size and quality of the interfaces, and the scoring function. We achieved some significant improvement compared to previous works, and a very high discriminative power on some specific functional classes. We provide a readout of the contributions of shape and physico-chemical complementarity, interface matching, and specificity, in the predictions. In addition, we assessed the ability of the approach to account for protein surface multiple usages, and we compared it with a sequence-based deep learning method. This work may contribute to guiding the exploitation of the large amounts of protein structural models now available toward the discovery of unexpected partners and their complex structure characterisation.

Identification of potential biomarkers in hepatocellular carcinoma: A network-based approach

BackgroundHepatocellular carcinoma (HCC) is one of the leading causes of death worldwide. Identification of potential therapeutic and diagnostic biomarkers can be helpful to screen cancer progress. This study was implemented to discover potential biomarkers for HCC within a network-based approach integrated with microarray data. MethodsThrough downloading a gene expression profile GSE62232 differentially expressed genes (DEGs) were identified. Gene ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis for DEGs were performed utilizing Enrichr server. Following reconstruction of protein-protein interaction network of DEGs with STRING, network visualization, analyses, and clustering into structural modules were carried out using Cytoscape. Considering degree centrality, 15 hub genes were selected as early biomarker candidates for final validation. To validate hub genes, the GEPIA server was used to perform overall survival (OS) and disease-free survival (DFS). ResultsIn our approach, 1996 DEGs were identified including 995 up-regulated genes and 1001 down-regulated genes. KEGG pathway enrichment analysis shows that DEGs are associated with chemical carcinogenesis and cell cycle. GO term enrichment analysis indicated the relation of DEGs with epoxygenase P450 pathway, arachidonic acid monooxygenase activity, and secretory granule lumen. Following analysis of the protein-protein interaction network of DEGs, the top three structural modules and 15 early hub genes were selected. Validation of hub genes was performed using GEPIA. Consequently, CDK1, CCNB1, CCNA2, CDC20, AURKA, MAD2L1, TOP2A, KIF11, BUB1B, TYMS, EZH2, and BUB1 were considered as our final proposed biomarkers. Conclusionusing an integrated network-based approach with microarray data our results revealed 12 final candidates with the potential to be considered as biomarkers in hepatocellular carcinoma.

Core genes involved in the regulation of acute lung injury and their association with COVID-19 and tumor progression: A bioinformatics and experimental study.

Acute lung injury (ALI) is a specific form of lung damage caused by different infectious and non-infectious agents, including SARS-CoV-2, leading to severe respiratory and systemic inflammation. To gain deeper insight into the molecular mechanisms behind ALI and to identify core elements of the regulatory network associated with this pathology, key genes involved in the regulation of the acute lung inflammatory response (Il6, Ccl2, Cat, Serpine1, Eln, Timp1, Ptx3, Socs3) were revealed using comprehensive bioinformatics analysis of whole-genome microarray datasets, functional annotation of differentially expressed genes (DEGs), reconstruction of protein-protein interaction networks and text mining. The bioinformatics data were validated using a murine model of LPS-induced ALI; changes in the gene expression patterns were assessed during ALI progression and prevention by anti-inflammatory therapy with dexamethasone and the semisynthetic triterpenoid soloxolone methyl (SM), two agents with different mechanisms of action. Analysis showed that 7 of 8 revealed ALI-related genes were susceptible to LPS challenge (up-regulation: Il6, Ccl2, Cat, Serpine1, Eln, Timp1, Socs3; down-regulation: Cat) and their expression was reversed by the pre-treatment of mice with both anti-inflammatory agents. Furthermore, ALI-associated nodal genes were analysed with respect to SARS-CoV-2 infection and lung cancers. The overlap with DEGs identified in postmortem lung tissues from COVID-19 patients revealed genes (Saa1, Rsad2, Ifi44, Rtp4, Mmp8) that (a) showed a high degree centrality in the COVID-19-related regulatory network, (b) were up-regulated in murine lungs after LPS administration, and (c) were susceptible to anti-inflammatory therapy. Analysis of ALI-associated key genes using The Cancer Genome Atlas showed their correlation with poor survival in patients with lung neoplasias (Ptx3, Timp1, Serpine1, Plaur). Taken together, a number of key genes playing a core function in the regulation of lung inflammation were found, which can serve both as promising therapeutic targets and molecular markers to control lung ailments, including COVID-19-associated ALI.

Evaluation of cardiovascular system state by urine proteome after manned space flight

In order to find markers to assess the functional state of the cardiovascular system (CVS) before and after spaceflight (first and seventh day after landing), we analyzed the urine proteome composition of 10 Russian cosmonauts aged of 35–51 years who have completed 169–199-day spaceflight onboard the International Space Station (ISS). Also an analysis of urine samples of 6 cosmonaut back-ups was conducted. A special sample preparation was performed, followed by liquid chromatography-mass spectrometry. It was shown that after long duration space flight concentration of several proteins of CVS in urine samples varies significantly. The dynamic of presence in urine of thioredoxin and apolipoprotein A-I could be related to the spaceflight, as it were not found before flight and in back-up controls, but were detected after the spaceflight. It was found that changes in cosmonauts’ urine proteome comprehensively reflect the adaptive responses of cardiovascular, renal and neuroendocrine systems to long-duration microgravity conditions.The use of bioinformatics analysis to the reconstruction of protein-protein interaction networks and the identification of overrepresented Gene Ontology (GO) biological processes related to the cardiovascular system allowed us to establish relationships between proteomic data and physiological effects observed in cosmonauts after the flight. Hypotheses on the possible pathogenesis and etiological factors causing this adaptive response were suggested in this work.

Protein-protein interaction sites prediction by ensemble random forests with synthetic minority oversampling technique.

The prediction of protein-protein interaction (PPI) sites is a key to mutation design, catalytic reaction and the reconstruction of PPI networks. It is a challenging task considering the significant abundant sequences and the imbalance issue in samples. A new ensemble learning-based method, Ensemble Learning of synthetic minority oversampling technique (SMOTE) for Unbalancing samples and RF algorithm (EL-SMURF), was proposed for PPI sites prediction in this study. The sequence profile feature and the residue evolution rates were combined for feature extraction of neighboring residues using a sliding window, and the SMOTE was applied to oversample interface residues in the feature space for the imbalance problem. The Multi-dimensional Scaling feature selection method was implemented to reduce feature redundancy and subset selection. Finally, the Random Forest classifiers were applied to build the ensemble learning model, and the optimal feature vectors were inserted into EL-SMURF to predict PPI sites. The performance validation of EL-SMURF on two independent validation datasets showed 77.1% and 77.7% accuracy, which were 6.2-15.7% and 6.1-18.9% higher than the other existing tools, respectively. The source codes and data used in this study are publicly available at http://github.com/QUST-AIBBDRC/EL-SMURF/. Supplementary data are available at Bioinformatics online.

Reconstruction of protein-protein interaction network of insulin signaling in Homo sapiens.

Diabetes is one of the most prevalent diseases in the world. Type 1 diabetes is characterized by the failure of synthesizing and secreting of insulin because of destroyed pancreatic β-cells. Type 2 diabetes, on the other hand, is described by the decreased synthesis and secretion of insulin because of the defect in pancreatic β-cells as well as by the failure of responding to insulin because of malfunctioning of insulin signaling. In order to understand the signaling mechanisms of responding to insulin, it is necessary to identify all components in the insulin signaling network. Here, an interaction network consisting of proteins that have statistically high probability of being biologically related to insulin signaling in Homo sapiens was reconstructed by integrating Gene Ontology (GO) annotations and interactome data. Furthermore, within this reconstructed network, interacting proteins which mediate the signal from insulin hormone to glucose transportation were identified using linear paths. The identification of key components functioning in insulin action on glucose metabolism is crucial for the efforts of preventing and treating type 2 diabetes mellitus.

Supervised reconstruction of biological networks with local models

Inference and reconstruction of biological networks from heterogeneous data is currently an active research subject with several important applications in systems biology. The problem has been attacked from many different points of view with varying degrees of success. In particular, predicting new edges with a reasonable false discovery rate is highly demanded for practical applications, but remains extremely challenging due to the sparsity of the networks of interest. While most previous approaches based on the partial knowledge of the network to be inferred build global models to predict new edges over the network, we introduce here a novel method which predicts whether there is an edge from a newly added vertex to each of the vertices of a known network using local models. This involves learning individually a certain subnetwork associated with each vertex of the known network, then using the discovered classification rule associated with only that vertex to predict the edge to the new vertex. Excellent experimental results are shown in the case of metabolic and protein-protein interaction network reconstruction from a variety of genomic data. An implementation of the proposed algorithm is available upon request from the authors.