Sophia Kossida

Using graph theory to analyze biological networks

Understanding complex systems often requires a bottom-up analysis towards a systems biology approach. The need to investigate a system, not only as individual components but as a whole, emerges. This can be done by examining the elementary constituents individually and then how these are connected. The myriad components of a system and their interactions are best characterized as networks and they are mainly represented as graphs where thousands of nodes are connected with thousands of vertices. In this article we demonstrate approaches, models and methods from the graph theory universe and we discuss ways in which they can be used to reveal hidden properties and features of a network. This network profiling combined with knowledge extraction will help us to better understand the biological significance of the system.

Search for potential markers for prostate cancer diagnosis, prognosis and treatment in clinical tissue specimens using amine-specific isobaric tagging (iTRAQ) with two-dimensional liquid chromatography and tandem mass spectrometry

This study aimed to identify candidate new diagnosis and prognosis markers and medicinal targets of prostate cancer (PCa), using state of the art proteomics. A total of 20 prostate tissue specimens from 10 patients with benign prostatic hyperplasia (BPH) and 10 with PCa (Tumour Node Metastasis [TNM] stage T1-T3) were analyzed by isobaric stable isotope labeling (iTRAQ) and two-dimensional liquid chromatography-tandem mass spectrometry (2DLC-MS/MS) approaches using a hybrid quadrupole time-of-flight system (QqTOF). The study resulted in the reproducible identification of 825 nonredundant gene products (p < or = 0.05) of which 30 exhibited up-regulation (> or =2-fold) and another 35 exhibited down-regulation (< or =0.5-fold) between the BPH and PCa specimens constituting a major contribution toward their global proteomic assessment. Selected findings were confirmed by immunohistochemical analysis of prostate tissue specimens. The proteins determined support existing knowledge and uncover novel and promising PCa biomarkers. The PCa proteome found can serve as a useful aid for the identification of improved diagnostic and prognostic markers and ultimately novel chemopreventive and therapeutic targets.

Human microRNA target analysis and gene ontology clustering by GOmir, a novel stand-alone application.

BackgroundmicroRNAs (miRNAs) are single-stranded RNA molecules of about 20–23 nucleotides length found in a wide variety of organisms. miRNAs regulate gene expression, by interacting with target mRNAs at specific sites in order to induce cleavage of the message or inhibit translation. Predicting or verifying mRNA targets of specific miRNAs is a difficult process of great importance.ResultsGOmir is a novel stand-alone application consisting of two separate tools: JTarget and TAGGO. JTarget integrates miRNA target prediction and functional analysis by combining the predicted target genes from TargetScan, miRanda, RNAhybrid and PicTar computational tools as well as the experimentally supported targets from TarBase and also providing a full gene description and functional analysis for each target gene. On the other hand, TAGGO application is designed to automatically group gene ontology annotations, taking advantage of the Gene Ontology (GO), in order to extract the main attributes of sets of proteins. GOmir represents a new tool incorporating two separate Java applications integrated into one stand-alone Java application.ConclusionGOmir (by using up to five different databases) introduces miRNA predicted targets accompanied by (a) full gene description, (b) functional analysis and (c) detailed gene ontology clustering. Additionally, a reverse search initiated by a potential target can also be conducted. GOmir can freely be downloaded BRFAA.

Evaluation of the Zucker Diabetic Fatty (ZDF) Rat as a Model for Human Disease Based on Urinary Peptidomic Profiles

Representative animal models for diabetes-associated vascular complications are extremely relevant in assessing potential therapeutic drugs. While several rodent models for type 2 diabetes (T2D) are available, their relevance in recapitulating renal and cardiovascular features of diabetes in man is not entirely clear. Here we evaluate at the molecular level the similarity between Zucker diabetic fatty (ZDF) rats, as a model of T2D-associated vascular complications, and human disease by urinary proteome analysis. Urine analysis of ZDF rats at early and late stages of disease compared to age- matched LEAN rats identified 180 peptides as potentially associated with diabetes complications. Overlaps with human chronic kidney disease (CKD) and cardiovascular disease (CVD) biomarkers were observed, corresponding to proteins marking kidney damage (eg albumin, alpha-1 antitrypsin) or related to disease development (collagen). Concordance in regulation of these peptides in rats versus humans was more pronounced in the CVD compared to the CKD panels. In addition, disease-associated predicted protease activities in ZDF rats showed higher similarities to the predicted activities in human CVD. Based on urinary peptidomic analysis, the ZDF rat model displays similarity to human CVD but might not be the most appropriate model to display human CKD on a molecular level.

Analysis of secreted proteins for the study of bladder cancer cell aggressiveness.

Secreted proteins play a key role in cell signaling, communication, and migration. We recently described the development of an aggressive variant (T24M) of the bladder cancer cell line T24. Using this cell line model, the objective of our work was the identification of secreted proteins involved in the acquisition of the aggressive phenotype. Using in vitro assays, we demonstrate that conditioned media of the T24M cells promote motility of the parental less aggressive T24 cells. Proteomic analysis of cell culture conditioned media by the use of 2-dimensional gel electrophoresis coupled to MALDI TOF MS and LC-MS approaches resulted in enrichment and detection of multiple classical extracellular and secreted proteins such as fibronectin, cystatin, fibrillin, fibulin, interleukin 6, etc. Comparison of the secretome of the T24 and T24M cells indicated differences in proteins with potential involvement in the mechanisms of cell aggressiveness including SPARC, tPA, and clusterin. These findings were further confirmed by Western blot analysis. In the case of SPARC, further studies involving transwell assays indicated that blockage of the protein in the presence of SPARC-specific Abs results in decreased cell motility. Collectively, our study provides a 2DE-based comprehensive analysis of bladder cancer cell secretome. The results indicate various secreted proteins with potential involvement in bladder cancer cell aggressiveness and more specifically provide initial evidence for special role of SPARC in bladder cancer cell motility and invasiveness.

Which clustering algorithm is better for predicting protein complexes

BackgroundProtein-Protein interactions (PPI) play a key role in determining the outcome of most cellular processes. The correct identification and characterization of protein interactions and the networks, which they comprise, is critical for understanding the molecular mechanisms within the cell. Large-scale techniques such as pull down assays and tandem affinity purification are used in order to detect protein interactions in an organism. Today, relatively new high-throughput methods like yeast two hybrid, mass spectrometry, microarrays, and phage display are also used to reveal protein interaction networks.ResultsIn this paper we evaluated four different clustering algorithms using six different interaction datasets. We parameterized the MCL, Spectral, RNSC and Affinity Propagation algorithms and applied them to six PPI datasets produced experimentally by Yeast 2 Hybrid (Y2H) and Tandem Affinity Purification (TAP) methods. The predicted clusters, so called protein complexes, were then compared and benchmarked with already known complexes stored in published databases.ConclusionsWhile results may differ upon parameterization, the MCL and RNSC algorithms seem to be more promising and more accurate at predicting PPI complexes. Moreover, they predict more complexes than other reviewed algorithms in absolute numbers. On the other hand the spectral clustering algorithm achieves the highest valid prediction rate in our experiments. However, it is nearly always outperformed by both RNSC and MCL in terms of the geometrical accuracy while it generates the fewest valid clusters than any other reviewed algorithm. This article demonstrates various metrics to evaluate the accuracy of such predictions as they are presented in the text below. Supplementary material can be found at: http://www.bioacademy.gr/bioinformatics/projects/ppireview.htm

State-of-the-art technology in modern computer-aided drug design

The quest for small drug-like compounds that selectively inhibit the function of biological targets has always been a major focus in the pharmaceutical industry and in academia as well. High-throughput screening of compound libraries requires time, cost and resources. Therefore, the use of alternative methods is necessary for facilitating lead discovery. Computational techniques that dock small molecules into macromolecular targets and predict the affinity and activity of the small molecule are widely used in drug design and discovery, and have become an integral part of the industrial and academic research. In this review, we present an overview of some state-of-the-art technologies in modern drug design that have been developed for expediting the search for novel drug candidates.

Phylogenetic analysis of the eukaryotic RNA (cytosine-5)-methyltransferases

RNA (cytosine-5)-methyltransferases (RCMTs) have been characterized both in prokaryotic and eukaryotic organisms. The RCMT family, however, remains largely uncharacterized, as opposed to the family of DNA (cytosine-5)-methyltransferases which has been studied in depth. In the present study, an in silico identification of the putative 5-methylcytosine RNA-generating enzymes in the eukaryotic genomes was performed. A comprehensive phylogenetic analysis of the putative eukaryotic RCMT-related proteins has been performed in order to redefine subfamilies within the RCMT family. Five distinct eukaryotic subfamilies were identified, including the three already known (NOP2, NCL1 and YNL022c), one novel subfamily (RCMT9) and a fifth one which hitherto was considered to exist exclusively in prokaryotes (Fmu). The potential evolutionary relationships among the different eukaryotic RCMT subfamilies were also investigated. Furthermore, the results of this study add further support to a previous hypothesis that RCMTs represent evolutionary intermediates of RNA (uridine-5)-methyltransferases and DNA (cytosine-5)-methyltransferases.

jClust: a clustering and visualization toolbox

jClust is a user-friendly application which provides access to a set of widely used clustering and clique finding algorithms. The toolbox allows a range of filtering procedures to be applied and is combined with an advanced implementation of the Medusa interactive visualization module. These implemented algorithms are k-Means, Affinity propagation, Bron–Kerbosch, MULIC, Restricted neighborhood search cluster algorithm, Markov clustering and Spectral clustering, while the supported filtering procedures are haircut, outside–inside, best neighbors and density control operations. The combination of a simple input file format, a set of clustering and filtering algorithms linked together with the visualization tool provides a powerful tool for data analysis and information extraction. Availability: http://jclust.embl.de/ Contact: pavlopou@embl.de; rschneid@embl.de; skossida@bioacademy.gr Supplementary information: Supplementary data are available at Bioinformatics online.

Noise reduction in protein-protein interaction graphs by the implementation of a novel weighting scheme

BackgroundRecent technological advances applied to biology such as yeast-two-hybrid, phage display and mass spectrometry have enabled us to create a detailed map of protein interaction networks. These interaction networks represent a rich, yet noisy, source of data that could be used to extract meaningful information, such as protein complexes. Several interaction network weighting schemes have been proposed so far in the literature in order to eliminate the noise inherent in interactome data. In this paper, we propose a novel weighting scheme and apply it to the S. cerevisiae interactome. Complex prediction rates are improved by up to 39%, depending on the clustering algorithm applied.ResultsWe adopt a two step procedure. During the first step, by applying both novel and well established protein-protein interaction (PPI) weighting methods, weights are introduced to the original interactome graph based on the confidence level that a given interaction is a true-positive one. The second step applies clustering using established algorithms in the field of graph theory, as well as two variations of Spectral clustering. The clustered interactome networks are also cross-validated against the confirmed protein complexes present in the MIPS database.ConclusionsThe results of our experimental work demonstrate that interactome graph weighting methods clearly improve the clustering results of several clustering algorithms. Moreover, our proposed weighting scheme outperforms other approaches of PPI graph weighting.