Methods for protein complex prediction and their contributions towards understanding the organisation, function and dynamics of complexes

BackgroundAccurate determination of protein complexes has become a key task of system biology for revealing cellular organization and function. Up to now, the protein complex prediction methods are mostly focused on static protein protein interaction (PPI) networks. However, cellular systems are highly dynamic and responsive to cues from the environment. The shift from static PPI networks to dynamic PPI networks is essential to accurately predict protein complex.ResultsThe gene expression data contains crucial dynamic information of proteins and PPIs, along with high-throughput experimental PPI data, are valuable for protein complex prediction. Firstly, we exploit gene expression data to calculate the active time point and the active probability of each protein and PPI. The dynamic active information is integrated into high-throughput PPI data to construct dynamic PPI networks. Secondly, a novel method for predicting protein complexes from the dynamic PPI networks is proposed based on core-attachment structural feature. Our method can effectively exploit not only the dynamic active information but also the topology structure information based on the dynamic PPI networks.ConclusionsWe construct four dynamic PPI networks, and accurately predict many well-characterized protein complexes. The experimental results show that (i) the dynamic active information significantly improves the performance of protein complex prediction; (ii) our method can effectively make good use of both the dynamic active information and the topology structure information of dynamic PPI networks to achieve state-of-the-art protein complex prediction capabilities.Electronic supplementary materialThe online version of this article (doi:10.1186/s12859-016-1101-y) contains supplementary material, which is available to authorized users.

Protein complex prediction based on maximum matching with domain–domain interaction

BackgroundProtein complexes play an important role in cellular mechanisms. Recently, several methods have been presented to predict protein complexes in a protein interaction network. In these methods, a protein complex is predicted as a dense subgraph of protein interactions. However, interactions data are incomplete and a protein complex does not have to be a complete or dense subgraph.ResultsWe propose a more appropriate protein complex prediction method, CFA, that is based on connectivity number on subgraphs. We evaluate CFA using several protein interaction networks on reference protein complexes in two benchmark data sets (MIPS and Aloy), containing 1142 and 61 known complexes respectively. We compare CFA to some existing protein complex prediction methods (CMC, MCL, PCP and RNSC) in terms of recall and precision. We show that CFA predicts more complexes correctly at a competitive level of precision.ConclusionsMany real complexes with different connectivity level in protein interaction network can be predicted based on connectivity number. Our CFA program and results are freely available from http://www.bioinf.cs.ipm.ir/softwares/cfa/CFA.rar.

Since the final decades of the last century, twin studies have made a remarkable contribution to the genetics of human complex traits and diseases. With the recent rapid development in modern biotechnology of high-throughput genetic and genomic analyses, twin modelling is expanding from analysis of diseases to molecular phenotypes in functional genomics especially in epigenetics, a thriving field of research that concerns the environmental regulation of gene expression through DNA methylation, histone modification, microRNA and long non-coding RNA expression, etc. The application of the twin method to molecular phenotypes offers new opportunities to study the genetic (nature) and environmental (nurture) contributions to epigenetic regulation of gene activity during developmental, ageing and disease processes. Besides the classical twin model, the case co-twin design using identical twins discordant for a trait or disease is becoming a popular and powerful design for epigenome-wide association study in linking environmental exposure to differential epigenetic regulation and to disease status while controlling for individual genetic make-up. It can be expected that novel uses of twin methods in epigenetic studies are going to help with efficiently unravelling the genetic and environmental basis of epigenomics in human complex diseases.

BackgroundSome proposed methods for identifying essential proteins have better results by using biological information. Gene expression data is generally used to identify essential proteins. However, gene expression data is prone to fluctuations, which may affect the accuracy of essential protein identification. Therefore, we propose an essential protein identification method based on gene expression and the PPI network data to calculate the similarity of "active" and "inactive" state of gene expression in a cluster of the PPI network. Our experiments show that the method can improve the accuracy in predicting essential proteins.ResultsIn this paper, we propose a new measure named JDC, which is based on the PPI network data and gene expression data. The JDC method offers a dynamic threshold method to binarize gene expression data. After that, it combines the degree centrality and Jaccard similarity index to calculate the JDC score for each protein in the PPI network. We benchmark the JDC method on four organisms respectively, and evaluate our method by using ROC analysis, modular analysis, jackknife analysis, overlapping analysis, top analysis, and accuracy analysis. The results show that the performance of JDC is better than DC, IC, EC, SC, BC, CC, NC, PeC, and WDC. We compare JDC with both NF-PIN and TS-PIN methods, which predict essential proteins through active PPI networks constructed from dynamic gene expression.ConclusionsWe demonstrate that the new centrality measure, JDC, is more efficient than state-of-the-art prediction methods with same input. The main ideas behind JDC are as follows: (1) Essential proteins are generally densely connected clusters in the PPI network. (2) Binarizing gene expression data can screen out fluctuations in gene expression profiles. (3) The essentiality of the protein depends on the similarity of "active" and "inactive" state of gene expression in a cluster of the PPI network.

The molecular dynamics of the inclusion complexes of cycloamyloses with phenylalanine in aqueous solution at different pD have been studied by means of carbon-13 NMR spectroscopy. As host cycloamyloses we have chosen cyclohexaamylose, cycloheptaamylose, and cyclooctaamylose. The influences of the cavity size of cycloamyloses and of medium pD values on the molecular dynamics of inclusion complexes were studied by using the NT1η values, where T1 is the carbon-13 spin-lattice relaxation time, N the number of directly bonded proton to a given carbon, and η is the viscosity of solution. Upon complexation, the NT1η values of phenyl carbons of guest phenylalanine show larger decreases than those of other carbons, indicating that the guest forms the inclusion complexes with the host by the insertion of its phenyl ring into the host cavity even in the case of cyclooctaamylose. It was found that the strength of the dynamic coupling between host and guest depends on the cavity size and pD values of medium. The strongest coupling was observed for the cycloheptaamylose–phenylalanine system at pD 11.3, where the phenyl ring is deeply and tightly included into the cavity. The inclusion of phenyl ring is shallow and loose in the case of cyclohexaamylose and deep and loose in the case of cyclooctaamylose.

Background: The wide use of genome wide association studies (GWAS) has led to the successful identification of multiple genetic susceptibility variants to several complex human diseases. Given the limited amount of data on genetic variation at these loci in populations of non-European origin, we investigated population variation among 11 population groups for loci showing strong and consistent association from GWAS with several complex human diseases. Methods: Data from the International HapMap Project Phase 3, comprising 11 population groups, were used to estimate allele frequencies at loci showing strong and consistent association from GWAS with any of 26 complex human diseases and traits. Allele frequency summary statistics and F_ST at each locus were used to estimate population differentiation. Results: There is wide variation in allele frequencies and F_ST across the 11 population groups for susceptibility loci to these complex human diseases and traits. Allele frequencies varied widely across populations, often by as much as 20- to 40-fold. F_ST, as a measure of population differentiation, also varied widely across the loci studied (for example, 0.019 to 0.201 for type 2 diabetes, 0.022 to 0.520 for prostate cancer loci, and 0.006 to 0.520 for serum lipid levels). Conclusions: The public health risk posed by any of these risk alleles is likely to show wide variation across populations simply as a function of its frequency, and this risk difference may be amplified by gene-gene and gene-environment interactions. These analyses offer compelling reasons for including multiple human populations from different parts of the world in the international effort to use genomic tools to understand disease etiology and differential distribution of diseases across ethnic groups.

Adenoid cystic carcinoma is a slow-growing malignancy that most often occurs in the salivary glands. Currently, no FDA-approved therapeutic target or diagnostic biomarker has been identified for this cancer. The aim of this study was to find new therapeutic and diagnostic targets using bioinformatics methods. We extracted the gene expression information from two GEO datasets (including GSE59701 and GSE88804). DEGs between ACC and normal samples were extracted using R software. The biochemical pathways involved in ACC were obtained by using the Enrichr database. PPI network was drawn by STRING, and important genes were extracted by Cytoscape. Real-time PCR and immunohistochemistry were used for biomarker verification. After analyzing the PPI network, 20 hub genes were introduced to have potential as diagnostic and therapeutic targets. Among these genes, PLCG1 was presented as new biomarker in ACC. Furthermore, by studying the function of the hub genes in the enriched biochemical pathways, we found that IGF-1R/IR and PPARG pathways most likely play a critical role in tumorigenesis and drug resistance in ACC and have a high potential for selection as therapeutic targets in future studies. In this study, we achieved the recognition of the pathways involving in ACC pathogenesis and also found potential targets for treatment and diagnosis of ACC. Further experimental studies are required to confirm the results of this study.

BackgroundMicroRNAs (miRNAs) have been confirmed to have close relationship with various human complex diseases. The identification of disease-related miRNAs provides great insights into the underlying pathogenesis of diseases. However, it is still a big challenge to identify which miRNAs are related to diseases. As experimental methods are in general expensive and time‐consuming, it is important to develop efficient computational models to discover potential miRNA-disease associations.MethodsThis study presents a novel prediction method called HFHLMDA, which is based on high-dimensionality features and hypergraph learning, to reveal the association between diseases and miRNAs. Firstly, the miRNA functional similarity and the disease semantic similarity are integrated to form an informative high-dimensionality feature vector. Then, a hypergraph is constructed by the K-Nearest-Neighbor (KNN) method, in which each miRNA-disease pair and its k most relevant neighbors are linked as one hyperedge to represent the complex relationships among miRNA-disease pairs. Finally, the hypergraph learning model is designed to learn the projection matrix which is used to calculate uncertain miRNA-disease association score.ResultCompared with four state-of-the-art computational models, HFHLMDA achieved best results of 92.09% and 91.87% in leave-one-out cross validation and fivefold cross validation, respectively. Moreover, in case studies on Esophageal neoplasms, Hepatocellular Carcinoma, Breast Neoplasms, 90%, 98%, and 96% of the top 50 predictions have been manually confirmed by previous experimental studies.ConclusionMiRNAs have complex connections with many human diseases. In this study, we proposed a novel computational model to predict the underlying miRNA-disease associations. All results show that the proposed method is effective for miRNA–disease association predication.

Enhancers, as specialized genomic cis-regulatory elements, activate transcription of their target genes and play an important role in pathogenesis of many human complex diseases. Despite recent systematic identification of them in the human genome, currently there is an urgent need for comprehensive annotation databases of human enhancers with a focus on their disease connections. In response, we built the Human Enhancer Disease Database (HEDD) to facilitate studies of enhancers and their potential roles in human complex diseases. HEDD currently provides comprehensive genomic information for ∼2.8 million human enhancers identified by ENCODE, FANTOM5 and RoadMap with disease association scores based on enhancer–gene and gene–disease connections. It also provides Web-based analytical tools to visualize enhancer networks and score enhancers given a set of selected genes in a specific gene network. HEDD is freely accessible at http://zdzlab.einstein.yu.edu/1/hedd.php.

Chapter 5 - Homology modeling: Developing 3D structures of target proteins missing in databases

Many functions in living cells are governed by macromolecular complexes. To fully describe the underlying mechanisms, they have to be understood at atomic level. The present study combines data obtained by X-ray crystallography and cryo-electron microscopy (cryo-EM) with molecular dynamics (MD) simulations. Two functions of macromolecular complexes, the downregulation of neurotransmitter release by the SNARE protein complex under oxidative stress and the translocation of transfer RNAs (tRNAs) through the ribosome during protein synthesis, were investigated. First, the hypothesis that oxidation of two cysteines on linker of the SNARE protein SNAP-25B and consequent disulfide bond formation shortens this linker sufficiently to hinder complex formation was tested. For this purpose, MD simulations of the SNARE complex with and without the disulfide bond were compared. Disulfide bond formation lead to conformational changes of the linker and of three central hydrophobic layers necessary to form the SNARE complex. Previously, mutations of residues contributing to these layers have been shown to reduce neurotransmitter release, suggesting that the stability of these layers is crucial for complex formation. The results from the simulations agree with the hypothesis that disulfide bond formation leads to a destabilization of the SNARE complex thus rendering it dysfunctional. This mechanism is interpreted as a chemomechanical regulation to shut down neurotransmitter release under oxidative stress, which has been linked to neurodegenerative diseases. In a second part I investigated the ribosome, where after peptide bond formation, bound tRNAs translocate by more than 7 nm to adjacent binding sites (A and P to P and E), accompanied by large-scale conformational motions (L1-stalk, intersubunit rotation) of the ribosome. By combining existing cryo-EM reconstructions of translocation intermediates with high resolution crystal structures, we obtained 13 near-atomic resolution structures. Subsequently, MD simulations of were carried out for each intermediate state. The obtained dynamics within these states allowed to estimate transition rates between states for motions of the L1-stalk, tRNAs and intersubunit rotations. Rapid motions of the L1-stalk and the small (30S) subunit on sub-microsecond timescales were revealed, whereas tRNA motions were seen to be rate-limiting for most transitions. By calculating the interaction free energy between L1-stalk and tRNA, molecular forces were derived showing that the L1-stalk is actively pulling the tRNA from P to E binding site, thereby overcoming barriers for the tRNA motion. Further, ribosomal proteins L5 and L16 guide the tRNAs by 'sliding' and 'stepping' mechanisms involving key protein-tRNA contacts. This explains how tRNA binding affinity is kept sufficiently constant to allow rapid translocation despite large-scale displacements. Translocation is accompanied by rotations of the 30S ribosomal subunit of more than 20 degrees relative to the large (50S) subunit. For each translocation intermediate, the affinity of the two subunits with each other must be finely tuned enabling such conformational flexibility while maintaining integrity of the ribosomal complex. Analyzing the trajectories at residue level reveals two classes of intersubunit contact interactions: i) persistent residue contacts which are independent of 30S rotation and primarily located close to the axis of rotation. ii) contacts that are formed and ruptured depending on the rotation angle, seen mainly on the periphery. Strikingly, also these rotation specific contacts substantially contribute to the overall stability of the ribosomal assembly and are expected to maintain a constant interaction energy with low barriers for rotation. The simulations reveal that upon removal of tRNAs peripheral contacts are weakened and, in turn, intersubunit rotation angles decrease, in agreement with cryo-EM analysis of tRNA depleted ribosomes. The identified mechanisms for lowering free energy barriers and for fine-tuning affinities might have developed similarly in other macromolecular complexes.

Mapping the pathogenic nexus: Gene overlap and protein interaction networks in Alzheimer's and breast cancer as a precursor to protein structure prediction and analysis.

BackgroundIn today’s society, cancer has become a big concern. The most common cancers in women are breast cancer (BC), endometrial cancer (EC), ovarian cancer (OC), and cervical cancer (CC). CC is a type of cervix cancer that is the fourth most common cancer in women and the fourth major cause of death. ResultsThis research uses a network approach to discover genetic connections, functional enrichment, pathways analysis, microRNAs transcription factors (miRNA-TF) co-regulatory network, gene-disease associations, and therapeutic targets for CC. Three datasets from the NCBI’s GEO collection were considered for this investigation. Then, using a comparison approach between the datasets, 315 common DEGs were discovered. The PPI network was built using a variety of combinatorial statistical approaches and bioinformatics tools, and the PPI network was then utilized to identify hub genes and critical modules. ConclusionFurthermore, we discovered that CC has specific similar links with the progression of different tumors using Gene Ontology terminology and pathway analysis. Transcription factors-gene linkages, gene-disease correlations, and the miRNA-TF co-regulatory network were revealed to have functional enrichments. We believe the candidate drugs identified in this study could be effective for advanced CC treatment.

Hepatocellular carcinoma (HCC) is still a significant global health problem. The development of bioinformatics may provide the opportunities to identify novel therapeutic targets. This study bioinformatically identified the differentially expressed genes (DEGs) in HCC and associated them with HCC prognosis using data from published databases. The DEGs downloaded from the Gene Expression Omnibus (GEO) website were visualized using the Venn diagram software, and then subjected to the GO and KEGG analyses, while the protein–protein interaction network was analyzed using Cytoscape software with the Search Tool for the search tool for the retrieval of interacting genes and the molecular complex detection plug-in. Kaplan–Meier curves and the log rank test were used to associate the core PPI network genes with the prognosis. There were 57 upregulated and 143 downregulated genes in HCC samples. The GO and pathway analyses revealed that these DEGs are involved in the biological processes (BPs), molecular functions (MFs), and cell components (CCs). The PPI network covered 50 upregulated and 108 downregulated genes, and the core modules of this PPI network contained 34 upregulated genes. A total of 28 of these upregulated genes were associated with a poor HCC prognosis, 27 of which were highly expressed in HCC tissues. This study identified 28 DEGs to be associated with a poor HCC prognosis. Future studies will investigate their possible applications as prognostic biomarkers and potential therapeutic targets for HCC.