Structure Of Protein-protein Interaction Networks Research Articles

Visualization of protein interaction networks: problems and solutions

BackgroundVisualization concerns the representation of data visually and is an important task in scientific research. Protein-protein interactions (PPI) are discovered using either wet lab techniques, such mass spectrometry, or in silico predictions tools, resulting in large collections of interactions stored in specialized databases. The set of all interactions of an organism forms a protein-protein interaction network (PIN) and is an important tool for studying the behaviour of the cell machinery. Since graphic representation of PINs may highlight important substructures, e.g. protein complexes, visualization is more and more used to study the underlying graph structure of PINs. Although graphs are well known data structures, there are different open problems regarding PINs visualization: the high number of nodes and connections, the heterogeneity of nodes (proteins) and edges (interactions), the possibility to annotate proteins and interactions with biological information extracted by ontologies (e.g. Gene Ontology) that enriches the PINs with semantic information, but complicates their visualization.MethodsIn these last years many software tools for the visualization of PINs have been developed. Initially thought for visualization only, some of them have been successively enriched with new functions for PPI data management and PIN analysis. The paper analyzes the main software tools for PINs visualization considering four main criteria: (i) technology, i.e. availability/license of the software and supported OS (Operating System) platforms; (ii) interoperability, i.e. ability to import/export networks in various formats, ability to export data in a graphic format, extensibility of the system, e.g. through plug-ins; (iii) visualization, i.e. supported layout and rendering algorithms and availability of parallel implementation; (iv) analysis, i.e. availability of network analysis functions, such as clustering or mining of the graph, and the possibility to interact with external databases.ResultsCurrently, many tools are available and it is not easy for the users choosing one of them. Some tools offer sophisticated 2D and 3D network visualization making available many layout algorithms, others tools are more data-oriented and support integration of interaction data coming from different sources and data annotation. Finally, some specialistic tools are dedicated to the analysis of pathways and cellular processes and are oriented toward systems biology studies, where the dynamic aspects of the processes being studied are central.ConclusionA current trend is the deployment of open, extensible visualization tools (e.g. Cytoscape), that may be incrementally enriched by the interactomics community with novel and more powerful functions for PIN analysis, through the development of plug-ins. On the other hand, another emerging trend regards the efficient and parallel implementation of the visualization engine that may provide high interactivity and near real-time response time, as in NAViGaTOR. From a technological point of view, open-source, free and extensible tools, like Cytoscape, guarantee a long term sustainability due to the largeness of the developers and users communities, and provide a great flexibility since new functions are continuously added by the developer community through new plug-ins, but the emerging parallel, often closed-source tools like NAViGaTOR, can offer near real-time response time also in the analysis of very huge PINs.

Connectedness of PPI network neighborhoods identifies regulatory hub proteins.

With the growing availability of high-throughput protein-protein interaction (PPI) data, it has become possible to consider how a protein's local or global network characteristics predict its function. We introduce a graph-theoretic approach that identifies key regulatory proteins in an organism by analyzing proteins' local PPI network structure. We apply the method to the yeast genome and describe several properties of the resulting set of regulatory hubs. Finally, we demonstrate how the identified hubs and putative target gene sets can be used to identify causative, functional regulators of differential gene expression linked to human disease. Code is available at http://bcb.cs.tufts.edu/hubcomps. fox.andrew.d@gmail.com; slonim@cs.tufts.edu Supplementary data are available at Bioinformatics online.

Resilience of protein–protein interaction networks as determined by their large-scale topological features

The relationship between the structure and function of biological networks constitutes a fundamental issue in systems biology. Particularly, the structure of protein-protein interaction networks is related to important biological functions. In this work, we investigated how such a resilience is determined by the large scale features of the respective networks. Four species are taken into account, namely yeast Saccharomyces cerevisiae, worm Caenorhabditis elegans, fly Drosophila melanogaster and Homo sapiens. We adopted two entropy-related measurements (degree entropy and dynamic entropy) in order to quantify the overall degree of robustness of these networks. We verified that while they exhibit similar structural variations under random node removal, they differ significantly when subjected to intentional attacks (hub removal). As a matter of fact, more complex species tended to exhibit more robust networks. More specifically, we quantified how six important measurements of the networks topology (namely clustering coefficient, average degree of neighbors, average shortest path length, diameter, assortativity coefficient, and slope of the power law degree distribution) correlated with the two entropy measurements. Our results revealed that the fraction of hubs and the average neighbor degree contribute significantly for the resilience of networks. In addition, the topological analysis of the removed hubs indicated that the presence of alternative paths between the proteins connected to hubs tend to reinforce resilience. The performed analysis helps to understand how resilience is underlain in networks and can be applied to the development of protein network models.

Difference in gene duplicability may explain the difference in overall structure of protein-protein interaction networks among eukaryotes

BackgroundA protein-protein interaction network (PIN) was suggested to be a disassortative network, in which interactions between high- and low-degree nodes are favored while hub-hub interactions are suppressed. It was postulated that a disassortative structure minimizes unfavorable cross-talks between different hub-centric functional modules and was positively selected in evolution. However, by re-examining yeast PIN data, several researchers reported that the disassortative structure observed in a PIN might be an experimental artifact. Therefore, the existence of a disassortative structure and its possible evolutionary mechanism remains unclear.ResultsIn this study, we investigated PINs from the yeast, worm, fly, human, and malaria parasite including four different yeast PIN datasets. The analyses showed that the yeast, worm, fly, and human PINs are disassortative while the malaria parasite PIN is not. By conducting simulation studies on the basis of a duplication-divergence model, we demonstrated that a preferential duplication of low- and high-degree nodes can generate disassortative and non-disassortative networks, respectively. From this observation, we hypothesized that the difference in degree dependence on gene duplications accounts for the difference in assortativity of PINs among species. Comparison of 55 proteomes in eukaryotes revealed that genes with lower degrees showed higher gene duplicabilities in the yeast, worm, and fly, while high-degree genes tend to have high duplicabilities in the malaria parasite, supporting the above hypothesis.ConclusionsThese results suggest that disassortative structures observed in PINs are merely a byproduct of preferential duplications of low-degree genes, which might be caused by an organism's living environment.

Diffusion Model Based Spectral Clustering for Protein-Protein Interaction Networks

BackgroundA goal of systems biology is to analyze large-scale molecular networks including gene expressions and protein-protein interactions, revealing the relationships between network structures and their biological functions. Dividing a protein-protein interaction (PPI) network into naturally grouped parts is an essential way to investigate the relationship between topology of networks and their functions. However, clear modular decomposition is often hard due to the heterogeneous or scale-free properties of PPI networks.Methodology/Principal FindingsTo address this problem, we propose a diffusion model-based spectral clustering algorithm, which analytically solves the cluster structure of PPI networks as a problem of random walks in the diffusion process in them. To cope with the heterogeneity of the networks, the power factor is introduced to adjust the diffusion matrix by weighting the transition (adjacency) matrix according to a node degree matrix. This algorithm is named adjustable diffusion matrix-based spectral clustering (ADMSC). To demonstrate the feasibility of ADMSC, we apply it to decomposition of a yeast PPI network, identifying biologically significant clusters with approximately equal size. Compared with other established algorithms, ADMSC facilitates clear and fast decomposition of PPI networks.Conclusions/SignificanceADMSC is proposed by introducing the power factor that adjusts the diffusion matrix to the heterogeneity of the PPI networks. ADMSC effectively partitions PPI networks into biologically significant clusters with almost equal sizes, while being very fast, robust and appealing simple.

Cover Picture: Proteomics 22'09

AbstractShigellosis, dysentery — by whatever name — Don't kiss the chimpsWith a 10–15% mortality rate when untreated in young children and immunedeficient patients, over one million deaths per year are attributable to this primate/human‐only disease. Furthermore, the Shigella spp. have developed resistance to b‐lactam, tetracycline, and aminoglycoside antibiotics very rapidly. A non‐primate model of the disease is based on gnotobiotic piglets. Aerobically grown bacteria shift to anaerobic metabolism when introduced into the GI tract of the piglets. Pieper et al. examined the aerobic and anaerobic proteomes using differential display methodology on 2‐D gels. Over 1050 separate gene products were identified. Bacterial defense genes and secretion system genes were among the types found. Novel proteins that might be targets for vaccines or drugs were also identified (OmpA, HtpG, and OspC2).Pieper, R. et al., Proteomics 2009, 9, 5029–5045.Form follow/leads functions? Essentiality, centrality and network topologyWho ever thought eight or ten years ago that we would be learning network‐speak to keep up with contemporary molecular and cell biology? Collaborating with mathematicians and physicists? Raise your hand. As I thought, not many. With the completion of the human genome DNA sequence it became clear that the sequence did not code for enough proteins to account for the complexity of life. At least two sources of complexity have been discovered to date: post‐translational modification and protein–protein interaction networks (PPINs). The structure of PPINs is examined in this work from Park et al., who raise the question (in Facebook terms): does popularity correlate with importance? They used two yeast PPINs to see how well centrality correlated with essentiality and found that, using 40 measures of centrality, the relationship was close for path‐based localized information centrality and gene essentiality. They found that random forest classifiers can work, too.Park, K. et al., Proteomics 2009, 9, 5143–5154.Standing on the shoulders of the great: Telomerase proteomicsIn those early days of molecular biology, when all “model replication systems” (bacteria and phages) went through circular intermediates, we did not need to worry about linear ends. Then we came to the end of the world, the end of the linear chromosome. Elizabeth Blackburn, Carol Greider, and Jack Szostak worked out the mechanics and enzymology of faithful replication of those linear ends and for that work received the 2009 Nobel Prize in physiology and medicine. Now researchers such as Zimmermann et al. can apply the tools of proteomics to tease apart the subtleties of telomere control of and control by other proteins. These researchers demonstrate the utility of SELDI MS/MS to examine the changes in specific cell fractions upon up‐ or down‐expression of telomere regulatory components. They confirmed several observations and found a new role for S100A6 (AKA “calcyclin”) in cell response to telomere dysfunction.Zimmermann, S. et al., Proteomics 2009, 9, 5175–5187.

Systems-level cancer gene identification from protein interaction network topology applied to melanogenesis-related functional genomics data

Many real-world phenomena have been described in terms of large networks. Networks have been invaluable models for the understanding of biological systems. Since proteins carry out most biological processes, we focus on analysing protein-protein interaction (PPI) networks. Proteins interact to perform a function. Thus, PPI networks reflect the interconnected nature of biological processes and analysing their structural properties could provide insights into biological function and disease. We have already demonstrated, by using a sensitive graph theoretic method for comparing topologies of node neighbourhoods called 'graphlet degree signatures', that proteins with similar surroundings in PPI networks tend to perform the same functions. Here, we explore whether the involvement of genes in cancer suggests the similarity of their topological 'signatures' as well. By applying a series of clustering methods to proteins' topological signature similarities, we demonstrate that the obtained clusters are significantly enriched with cancer genes. We apply this methodology to identify novel cancer gene candidates, validating 80 per cent of our predictions in the literature. We also validate predictions biologically by identifying cancer-related negative regulators of melanogenesis identified in our siRNA screen. This is encouraging, since we have done this solely from PPI network topology. We provide clear evidence that PPI network structure around cancer genes is different from the structure around non-cancer genes. Understanding the underlying principles of this phenomenon is an open question, with a potential for increasing our understanding of complex diseases.

Non-Uniform Survival Rate of Heterodimerization Links in the Evolution of the Yeast Protein-Protein Interaction Network

Protein-protein interaction networks (PINs) are scale-free networks with a small-world property. In a small-world network, the average cluster coefficient (<C>) is much higher than in a random network, but the average shortest path length (<L>) is similar between the two networks. To understand the evolutionary mechanisms shaping the structure of PINs, simulation studies using various network growth models have been performed. It has been reported that the heterodimerization (HD) model, in which a new link is added between duplicated nodes with a uniform probability, could reproduce scale-freeness and a high <C>. In this paper, however, we show that the HD model is unsatisfactory, because (i) to reproduce the high <C> in the yeast PIN, a much larger number (n HI) of HD links (links between duplicated nodes) are required than the estimated number of n HI in the yeast PIN and (ii) the spatial distribution of triangles in the yeast PIN is highly skewed but the HD model cannot reproduce the skewed distribution. To resolve these discrepancies, we here propose a new model named the non-uniform heterodimerization (NHD) model. In this model, an HD link is preferentially attached between duplicated nodes when they share many common neighbors. Simulation studies demonstrated that the NHD model can successfully reproduce the high <C>, the low n HI, and the skewed distribution of triangles in the yeast PIN. These results suggest that the survival rate of HD links is not uniform in the evolution of PINs, and that an HD link between high-degree nodes tends to be evolutionarily conservative. The non-uniform survival rate of HD links can be explained by assuming a low mutation rate for a high-degree node, and thus this model appears to be biologically plausible.

Finding molecular complexes through multiple layer clustering of protein interaction networks

Clustering protein-protein interaction networks (PINs) helps to identify complexes that guide the cell machinery. Clustering algorithms often create a flat clustering, without considering the layered structure of PINs. We propose the MULIC clustering algorithm that produces layered clusters. We applied MULIC to five PINs. Clusters correlate with known MIPS protein complexes. For example, a cluster of 79 proteins overlaps with a known complex of 88 proteins. Proteins in top cluster layers tend to be more representative of complexes than proteins in bottom layers. Lab work on finding unknown complexes or determining drug effects can be guided by top layer proteins.

Combining gene expression profiles and protein–protein interaction data to infer gene functions

The ever-increasing flow of gene expression profiles and protein-protein interactions has catalyzed many computational approaches for inference of gene functions. Despite all the efforts, there is still room for improvement, for the information enriched in each biological data source has not been exploited to its fullness. A composite method is proposed for classifying unannotated genes based on expression data and protein-protein interaction (PPI) data, which extracts information from both data sources in novel ways. With the noise nature of expression data taken into consideration, importance is attached to the consensus expression patterns of gene classes instead of the actual expression profiles of individual genes, thus characterizing the composite method with enhanced robustness against microarray data variation. With regard to the PPI network, the traditional clear-cut binary attitude towards inter- and intra-functional interactions is abandoned, whereas a more objective perspective into the PPI network structure is formed through incorporating the varied function-function interaction probabilities into the algorithm. The composite method was implemented in two numerical experiments, where its improvement over single-data-source based methods was observed and the superiority of the novel data handling operations was discussed.

On the structure of protein-protein interaction networks.

We present a simple model for the underlying structure of protein-protein pairwise interaction graphs that is based on the way in which proteins attach to each other in experiments such as yeast two-hybrid assays. We show that data on the interactions of human proteins lend support to this model. The frequency of the number of connections per protein under this model does not follow a power law, in contrast to the reported behaviour of data from large-scale yeast two-hybrid screens of yeast protein-protein interactions. Sampling sub-graphs from the underlying graphs generated with our model, in a way analogous to the sampling performed in large-scale yeast two-hybrid searches, gives degree distributions that differ subtly from the power law and that fit the observed data better than the power law itself. Our results show that the observation of approximate power law behaviour in a sampled sub-graph does not imply that the underlying graph follows a power law.