Molecular interaction networks for the analysis of human disease: Utility,limitations, and considerations

High‐throughput ‘‐omics’ data can be combined with large‐scale molecular interaction networks, for example, protein–protein interaction networks, to provide a unique framework for the investigation of human molecular biology. Interest in these integrative ‘‐omics’ methods is growing rapidly because of their potential to understand complexity and association with disease; such approaches have a focus on associations between phenotype and “network‐type.” The potential of this research is enticing, yet there remain a series of important considerations. Here, we discuss interaction data selection, data quality, the relative merits of using data from large high‐throughput studies versus a meta‐database of smaller literature‐curated studies, and possible issues of sociological or inspection bias in interaction data. Other work underway, especially international consortia to establish data formats, quality standards and address data redundancy, and the improvements these efforts are making to the field, is also evaluated. We present options for researchers intending to use large‐scale molecular interaction networks as a functional context for protein or gene expression data, including microRNAs, especially in the context of human disease.     

Identification of the interactome between fish plasma proteins and Edwardsiella tarda reveals tissue-specific strategies against bacterial infection

Elucidating the complex pathogen-host interaction is essential for a comprehensive understanding of how these remarkable agents invade their hosts and how the hosts defend against these invaders. During the infection, pathogens interact intensively with host to enable their survival, which can be revealed through their interactome. Edwardsiella tarda is a Gram-negative bacterial pathogen causing huge economic loss in aquaculture and a spectrum of intestinal and extraintestinal diseases in humans. E. tarda is an ideal model for host-pathogen investigation as it infects fish in three distinct steps: entering the host, circulating through the blood and establishing infection. We adopted a previous established proteomic approach that inactivated E. tarda cells and covalent crosslink fish plasma proteins were used to capture plasma proteins and bacterial outer membrane proteins, respectively. By the combinatorial use of proteomic and biochemical approaches, six plasma proteins and seven outer membrane proteins (OMPs) were identified. Interactions among these proteins were validated with protein-array, far-Western blotting and co-immunoprecipitation. At last, seventeen plasma protein-bacteria protein⿿protein interaction were confirmed to be involved in the interaction network, forming a complex interactome. Compared to our previous results, different host proteins were detected, whereas some of the bacterial proteins were similar, which indicates that hosts adopt tissue-specific strategies to cope with the same pathogen during infection. Thus, our results provide a robust demonstration of both bacterial initiators and host receptors or interacting proteins to further explore infection and anti-infective mechanisms between hosts and microbes.     

The protein–protein interaction network of the human Sirtuin family

Protein–protein interaction networks are useful for studying human diseases and to look for possible health care through a holistic approach. Networks are playing an increasing and important role in the understanding of physiological processes such as homeostasis, signaling, spatial and temporal organizations, and pathological conditions. In this article we show the complex system of interactions determined by human Sirtuins (Sirt) largely involved in many metabolic processes as well as in different diseases. The Sirtuin family consists of seven homologous Sirt-s having structurally similar cores but different terminal segments, being rather variable in length and/or intrinsically disordered. Many studies have determined their cellular location as well as biological functions although molecular mechanisms through which they act are actually little known therefore, the aim of this work was to define, explore and understand the Sirtuin-related human interactome. As a first step, we have integrated the experimentally determined protein–protein interactions of the Sirtuin-family as well as their first and second neighbors to a Sirtuin-related sub-interactome. Our data showed that the second-neighbor network of Sirtuins encompasses 25% of the entire human interactome, and exhibits a scale-free degree distribution and interconnectedness among top degree nodes. Moreover, the Sirtuin sub interactome showed a modular structure around the core comprising mixed functions. Finally, we extracted from the Sirtuin sub-interactome subnets related to cancer, aging and post-translational modifications for information on key nodes and topological space of the subnets in the Sirt family network.     

Kiwi: a tool for integration and visualization of network topology and gene-set analysis

Background

The analysis of high-throughput data in biology is aided by integrative approaches such as gene-set analysis. Gene-sets can represent well-defined biological entities (e.g. metabolites) that interact in networks (e.g. metabolic networks), to exert their function within the cell. Data interpretation can benefit from incorporating the underlying network, but there are currently no optimal methods that link gene-set analysis and network structures.

Results

Here we present Kiwi, a new tool that processes output data from gene-set analysis and integrates them with a network structure such that the inherent connectivity between gene-sets, i.e. not simply the gene overlap, becomes apparent. In two case studies, we demonstrate that standard gene-set analysis points at metabolites regulated in the interrogated condition. Nevertheless, only the integration of the interactions between these metabolites provides an extra layer of information that highlights how they are tightly connected in the metabolic network.

Conclusions

Kiwi is a tool that enhances interpretability of high-throughput data. It allows the users not only to discover a list of significant entities or processes as in gene-set analysis, but also to visualize whether these entities or processes are isolated or connected by means of their biological interaction. Kiwi is available as a Python package at http://www.sysbio.se/kiwi and an online tool in the BioMet Toolbox at http://www.biomet-toolbox.org.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-014-0408-9) contains supplementary material, which is available to authorized users.     

Global insights into protein complexes through integrated analysis of the reliable interactome and knockout lethality

We have investigated the novel function of intracellular reactive oxygen species (ROS) in the activation of in situ tissue transglutaminase (tTGase) by lysophosphatidic acid (LPA) and transforming growth factor-beta (TGF-beta) in Swiss 3T3 fibroblasts. LPA induced a transient increase of intracellular ROS with a maximal increase at 10 min, which was blocked by ROS scavengers, N-acetyl-L-cysteine and catalase. LPA activated tTGase with a maximal increase at 1h, which was inhibited by cystamine and ROS scavengers. Incubation with exogenous H(2)O(2) activated tTGase. TGF-beta also activated tTGase with a maximal activation at 2h and the tTGase activation was inhibited by the ROS scavengers. Scrape-loading of C3 transferase inhibited the ROS production and in situ tTGase activation by LPA and TGF-beta, and the inhibitory effect of C3 transferase was reversed by exogenous H(2)O(2). Microinjection of GTPgammaS inhibited transamidating activity of tTGase stimulated by LPA, TGF-beta, and maitotoxin. These results suggested that intracellular ROS was essential for the activation of in situ tTGase in response to LPA and TGF-beta.     

Computational analysis of human protein interaction networks

Large amounts of human protein interaction data have been produced by experiments and prediction methods. However, the experimental coverage of the human interactome is still low in contrast to predicted data. To gain insight into the value of publicly available human protein network data, we compared predicted datasets, high-throughput results from yeast two-hybrid screens, and literature-curated protein-protein interactions. This evaluation is not only important for further methodological improvements, but also for increasing the confidence in functional hypotheses derived from predictions. Therefore, we assessed the quality and the potential bias of the different datasets using functional similarity based on the Gene Ontology, structural iPfam domain-domain interactions, likelihood ratios, and topological network parameters. This analysis revealed major differences between predicted datasets, but some of them also scored at least as high as the experimental ones regarding multiple quality measures. Therefore, since only small pair wise overlap between most datasets is observed, they may be combined to enlarge the available human interactome data. For this purpose, we additionally studied the influence of protein length on data quality and the number of disease proteins covered by each dataset. We could further demonstrate that protein interactions predicted by more than one method achieve an elevated reliability.     

基于Cytoscape的蛋白质网络可视化聚类分析插件

蛋白质网络聚类是识别功能模块的重要手段,不仅有利于理解生物系统的组织结构,对预测蛋白质功能也具有重要的意义.聚类结果的可视化分析是实现蛋白质网络聚类的有效途径.本论文基于开源的Cytoscape平台,设计并实现了一个蛋白质网络聚类分析及可视化插件CytoCluster.该插件集成了MCODE,FAG-EC,HC-PIN,OH-PIN,IPCA,EAGLE等六种典型的聚类算法;实现了聚类结果的可视化,将分析所得的clusters以缩略图列表的形式直观地显示出来,对于单个cluster,可显示在原网络中的位置,并能生成相应的子图单独显示;可对聚类结果进行导出,记录了算法名称、参数、聚类结果等信息.该插件具有良好的扩展性,提供了统一的算法接口,可不断添加新的聚类算法.     

Efficient prediction of human protein-protein interactions at a global scale

Background

Our knowledge of global protein-protein interaction (PPI) networks in complex organisms such as humans is hindered by technical limitations of current methods.

Results

On the basis of short co-occurring polypeptide regions, we developed a tool called MP-PIPE capable of predicting a global human PPI network within 3 months. With a recall of 23% at a precision of 82.1%, we predicted 172,132 putative PPIs. We demonstrate the usefulness of these predictions through a range of experiments.

Conclusions

The speed and accuracy associated with MP-PIPE can make this a potential tool to study individual human PPI networks (from genomic sequences alone) for personalized medicine.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-014-0383-1) contains supplementary material, which is available to authorized users.     

Integration of omics sciences to advance biology and medicine

In the past two decades, our ability to study cellular and molecular systems has been transformed through the development of omics sciences. While unlimited potential lies within massive omics datasets, the success of omics sciences to further our understanding of human disease and/or translating these findings to clinical utility remains elusive due to a number of factors. A significant limiting factor is the integration of different omics datasets (i.e., integromics) for extraction of biological and clinical insights. To this end, the National Cancer Institute (NCI) and the National Heart, Lung and Blood Institute (NHLBI) organized a joint workshop in June 2012 with the focus on integration issues related to multi-omics technologies that needed to be resolved in order to realize the full utility of integrating omics datasets by providing a glimpse into the disease as an integrated “system”. The overarching goals were to (1) identify challenges and roadblocks in omics integration, and (2) facilitate the full maturation of ‘integromics’ in biology and medicine. Participants reached a consensus on the most significant barriers for integrating omics sciences and provided recommendations on viable approaches to overcome each of these barriers within the areas of technology, bioinformatics and clinical medicine.     

Proteome analysis of multidrug resistance in vincristine-resistant human gastric cancer cell line SGC7901/VCR

In order to elucidate the mechanisms of multidrug resistance (MDR) of vincristine-resistant human gastric carcinoma cell line SGC7901/VCR, 2-DE was used to separate the total proteins of SGC7901/VCR and its parental cell line SGC7901. PDQuest software was applied to analyze 2-DE images, and the differential protein spots were identified by both MALDI-TOF-MS and ESI-Q-TOF-MS. Then the differential expressional levels of partially identified proteins were determined by Western blot analysis and real-time RT-PCR. Furthermore, the association of heat shock protein (HSP27), one of the highly expressed proteins in sgc7901/vcr, with MDR was analyzed using antisense inhibition of HSP27. In this study, the well-resolved, reproducible 2-DE patterns of SGC7901/VCR and SGC7901 were established, and yielded about 1100 protein-spots each. All the 24 differential proteins between the two cell lines were identified, and the differential expression levels of the partial proteins were confirmed. The suppression of HSP27 expression by HSP27 antisense oligonucleotides could enhance vincristine chemosensitivity in sgc7901/vcr and induce the cells to exhibit apoptotic morphological features after vincristine treatment. The differentially expressed proteins could be divided into six groups based on their functions: calcium-binding proteins, chaperones, proteins involved in drug detoxification or repair of DNA damage, metabolic enzymes, proteins related to cellular structure, and proteins relative to signal transduction, some of which may contribute to MDR of human gastric carcinoma cell line SGC7901/VCR. These data will be valuable for further study of the mechanisms of MDR in human gastric cancer.     

The interactome of LIM domain proteins: the contributions of LIM domain proteins to heart failure and heart development

LIM domain proteins all contain at least one double zinc-finger motif. They belong to a large family and here we review those expressed mainly in mammalian hearts, but particularly in cardiomyocytes. These proteins contain between one and five LIM domains and usually these proteins contain other domains that have specific functions such as actin-binding, kinases and nuclear translocation motifs. While several recent reviews have summarised the importance of individual LIM domain proteins, this is the first review of its kind to cover all LIMs associated with the heart. Here we examine 33 LIM proteins (including three that bind to, but do not themselves contain, LIM domains) that are implicated in either the development of the heart, heart disorders and failure, or both. Our analysis is consistent with the view that cardiac LIM domain proteins form multiple extensive networks of multi-protein complexes within the myocardium. This multiplicity of binding partners probably protects the heart as it is challenged to maintain cardiac output, until the imbalance reaches a turning point that results in failure. We believe that the complexity of LIM interactions is properly described by the term LIM interactome.     

FunRich: An open access standalone functional enrichment and interaction network analysis tool

As high‐throughput techniques including proteomics become more accessible to individual laboratories, there is an urgent need for a user‐friendly bioinformatics analysis system. Here, we describe FunRich, an open access, standalone functional enrichment and network analysis tool. FunRich is designed to be used by biologists with minimal or no support from computational and database experts. Using FunRich, users can perform functional enrichment analysis on background databases that are integrated from heterogeneous genomic and proteomic resources (>1.5 million annotations). Besides default human specific FunRich database, users can download data from the UniProt database, which currently supports 20 different taxonomies against which enrichment analysis can be performed. Moreover, the users can build their own custom databases and perform the enrichment analysis irrespective of organism. In addition to proteomics datasets, the custom database allows for the tool to be used for genomics, lipidomics and metabolomics datasets. Thus, FunRich allows for complete database customization and thereby permits for the tool to be exploited as a skeleton for enrichment analysis irrespective of the data type or organism used. FunRich ( http://www.funrich.org ) is user‐friendly and provides graphical representation (Venn, pie charts, bar graphs, column, heatmap and doughnuts) of the data with customizable font, scale and color (publication quality).     

Application of Meta-Mesh on the analysis of microbial communities from human associated-habitats

With the current fast accumulation of microbial community samples and related metagenomic sequencing data, data integration and analysis system is urgently needed for in-depth analysis of large number of metagenomic samples (also referred to as “microbial communities”) of interest. Although several existing databases have collected a large number of metagenomic samples, they mostly serve as data repositories with crude annotations, and offer limited functionality for analysis. Moreover, the few available tools for comparative analysis in the literature could only support the comparison of a few pre-defined set of metagenomic samples. To facilitate comprehensive comparative analysis on large amount of diverse microbial community samples, we have designed a Meta-Mesh system for a variety of analyses including quantitative analysis of similarities among microbial communities and computation of the correlation between the meta-information of these samples. We have used Meta-Mesh for systematically and efficiently analyses on diverse sets of human associate-habitat microbial community samples. Results have shown that Meta-Mesh could serve well as an efficient data analysis platform for discovery of clusters, biomarker and other valuable biological information from a large pool of human microbial samples.     

Protein interactome analysis of 12 mitogen‐activated protein kinase kinase kinase in rice using a yeast two‐hybrid system

The mitogen‐activated protein kinase (MAPK) cascade is composed at least of MAP3K (for MAPK kinase kinase), MAP2K, and MAPK family modules. These components together play a central role in mediating extracellular signals to the cell and vice versa by interacting with their partner proteins. However, the MAP3K‐interacting proteins remain poorly investigated in plants. Here, we utilized a yeast two‐hybrid system and bimolecular fluorescence complementation in the model crop rice (Oryza sativa) to map MAP3K‐interacting proteins. We identified 12 novel nonredundant interacting protein pairs (IPPs) representing 11 nonredundant interactors using 12 rice MAP3Ks (available as full‐length cDNA in the rice KOME ( http://cdna01.dna.affrc.go.jp/cDNA/ ) at the time of experimental design and execution) as bait and a rice seedling cDNA library as prey. Of the 12 MAP3Ks, only six had interacting protein partners. The established MAP3K interactome consisted of two kinases, three proteases, two forkhead‐associated domain‐containing proteins, two expressed proteins, one E3 ligase, one regulatory protein, and one retrotransposon protein. Notably, no MAP3K showed physical interaction with either MAP2K or MAPK. Seven IPPs (58.3%) were confirmed in vivo by bimolecular fluorescence complementation. Subcellular localization of 14 interactors, together involved in nine IPPs (75%) further provide prerequisite for biological significance of the IPPs. Furthermore, GO of identified interactors predicted their involvement in diverse physiological responses, which were supported by a literature survey. These findings increase our knowledge of the MAP3K‐interacting proteins, help in proposing a model of MAPK modules, provide a valuable resource for developing a complete map of the rice MAPK interactome, and allow discussion for translating the interactome knowledge to rice crop improvement against environmental factors.     

A comprehensive in silico analysis of huntingtin and its interactome

A polyglutamine expansion of the N-terminal region of huntingtin (Htt) causes Huntington’s disease, a severe neurodegenerative disorder. Htt huge multidomain structure, the presence of disordered regions, and the lack of sequence homologs of known structure, so far prevented structural studies of Htt, making the study of its structure-function relationships very difficult. In this work, the presence and location of five Htt ordered domains (named from Hunt1 to Hunt5) has been detected and the structure of these domains has been predicted for the first time using a combined threading/ab initio modeling approach. This work has led to the identification of a previously undetected HEAT repeats region in the Hunt3 domain. Furthermore, a putative function has been assigned to four out of the five domains. Hunt1 and Hunt5, displaying structural similarity with the regulatory subunit A of protein phosphatase 2A, are predicted to play a role in regulating the phosphorylation status of cellular proteins. Hunt2 and Hunt3 are predicted to be homologs of two yeast importins and to mediate vescicles transport and protein trafficking. Finally, a comprehensive analysis of the Htt interactome has been carried out and is discussed to provide a global picture of the Htt’s structure–function relationships.     

Ovarian cancer detection by logical analysis of proteomic data

A new type of efficient and accurate proteomic ovarian cancer diagnosis systems is proposed. The system is developed using the combinatorics and optimization-based methodology of logical analysis of data (LAD) to the Ovarian Dataset 8-7-02 (http://clinicalproteomics.steem.com), which updates the one used by Petricoin et al. in The Lancet 2002, 359, 572-577. This mass spectroscopy-generated dataset contains expression profiles of 15 154 peptides defined by their mass/charge ratios (m/z) in serum of 162 ovarian cancer and 91 control cases. Several fully reproducible models using only 7-9 of the 15 154 peptides were constructed, and shown in multiple cross-validation tests (k-folding and leave-one-out) to provide sensitivities and specificities of up to 100%. A special diagnostic system for stage I ovarian cancer patients is shown to have similarly high accuracy. Other results: (i) expressions of peptides with relatively low m/z values in the dataset are shown to be better at distinguishing ovarian cancer cases from controls than those with higher m/z values; (ii) two large groups of patients with a high degree of similarities among their formal (mathematical) profiles are detected; (iii) several peptides with a blocking or promoting effect on ovarian cancer are identified.     

TSUNAMI: Translational Bioinformatics Tool Suite for Network Analysis and Mining

Gene co-expression network (GCN) mining identifies gene modules with highly correlated expression profiles across samples/conditions. It enables researchers to discover latent gene/molecule interactions, identify novel gene functions, and extract molecular features from certain disease/condition groups, thus helping to identify disease biomarkers. However, there lacks an easy-to-use tool package for users to mine GCN modules that are relatively small in size with tightly connected genes that can be convenient for downstream gene set enrichment analysis, as well as modules that may share common members. To address this need, we developed an online GCN mining tool package: TSUNAMI (Tools SUite for Network Analysis and MIning). TSUNAMI incorporates our state-of-the-art lmQCM algorithm to mine GCN modules for both public and user-input data (microarray, RNA-seq, or any other numerical omics data), and then performs downstream gene set enrichment analysis for the identified modules. It has several features and advantages: 1) a user-friendly interface and real-time co-expression network mining through a web server; 2) direct access and search of NCBI Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) databases, as well as user-input gene expression matrices for GCN module mining; 3) multiple co-expression analysis tools to choose from, all of which are highly flexible in regards to parameter selection options; 4) identified GCN modules are summarized to eigengenes, which are convenient for users to check their correlation with other clinical traits; 5) integrated downstream Enrichr enrichment analysis and links to other gene set enrichment tools; and 6) visualization of gene loci by Circos plot in any step of the process. The web service is freely accessible through URL: https://biolearns.medicine.iu.edu/. Source code is available at https://github.com/huangzhii/TSUNAMI/.