Ioannis Kanaris

WORDS VERSUS CHARACTER N-GRAMS FOR ANTI-SPAM FILTERING

The increasing number of unsolicited e-mail messages (spam) reveals the need for the development of reliable anti-spam filters. The vast majority of content-based techniques rely on word-based repr...

Webpage Genre Identification Using Variable-Length Character n-Grams

An important factor for discriminating between Web pages is their genre (e.g., blogs, personal homepages, e-shops, online newspapers, etc). Web page genre identification has a great potential in information retrieval since users of search engines can combine genre-based and traditional topic-based queries to improve the quality of the results. So far, various features have been proposed to quantify the style of Web pages including word and HTML-tag frequencies. In this paper, we propose a low-level representation for this problem based on character n-grams. Using an existing approach, we produce feature sets of variable-length character n- grams and combine this representation with information about the most frequent HTML-tags. Based on two benchmark corpora, we present Web page genre identification experiments and improve the best reported results in both cases.

KEGGconverter: a tool for the in-silico modelling of metabolic networks of the KEGG Pathways database

BackgroundThe KEGG Pathway database is a valuable collection of metabolic pathway maps. Nevertheless, the production of simulation capable metabolic networks from KEGG Pathway data is a challenging complicated work, regardless the already developed tools for this scope. Originally used for illustration purposes, KEGG Pathways through KGML (KEGG Markup Language) files, can provide complete reaction sets and introduce species versioning, which offers advantages for the scope of cellular metabolism simulation modelling. In this project, KEGGconverter is described, implemented also as a web-based application, which uses as source KGML files, in order to construct integrated pathway SBML models fully functional for simulation purposes.ResultsA case study of the integration of six human metabolic pathways from KEGG depicts the ability of KEGGconverter to automatically produce merged and converted to SBML fully functional pathway models, enhanced with default kinetics. The suitability of the developed tool is demonstrated through a comparison with other state-of-the art relevant software tools for the same data fusion and conversion tasks, thus illustrating the problems and the relevant workflows. Moreover, KEGGconverter permits the inclusion of additional reactions in the resulting model which represent flux cross-talk with neighbouring pathways, providing in this way improved simulative accuracy. These additional reactions are introduced by exploiting relevant semantic information for the elements of the KEGG Pathways database. The architecture and functionalities of the web-based application are presented.ConclusionKEGGconverter is capable of producing integrated analogues of metabolic pathways appropriate for simulation tasks, by inputting only KGML files. The web application acts as a user friendly shell which transparently enables the automated biochemically correct pathway merging, conversion to SBML format, proper renaming of the species, and insertion of default kinetic properties for the pertaining reactions. The tool is available at: http://www.grissom.gr/keggconverter

Learning to recognize webpage genres

Webpages are mainly distinguished by their topic (e.g., politics, sports etc.) and genre (e.g., blogs, homepages, e-shops, etc.). Automatic detection of webpage genre could considerably enhance the ability of modern search engines to focus on the requirements of the users information need. In this paper, we present an approach to webpage genre detection based on a fully-automated extraction of the feature set that represents the style of webpages. The features we propose (character n-grams of variable length and HTML tags) are language-independent and easily-extracted while they can be adapted to the properties of the still evolving web genres and the noisy environment of the web. Experiments based on two publicly-available corpora show that the performance of the proposed approach is superior in comparison to previously reported results. It is also shown that character n-grams are better features than words when the dimensionality increases while the binary representation is more effective than the term-frequency representation for both feature types. Moreover, we perform a series of cross-check experiments (e.g., training using a genre palette and testing using a different genre palette as well as using the features extracted from one corpus to discriminate the genres of the other corpus) to illustrate the robustness of our approach and its ability to capture the general stylistic properties of genre categories even when the feature set is not optimized for the given corpus.

Stability of feature selection algorithms for classification in high-throughput genomics datasets

A major goal of the application of Machine Learning techniques to high-throughput genomics data (e.g. DNA microarrays or RNA-Seq), is the identification of “gene signatures”. These signatures can be used to discriminate among healthy or disease states (e.g. normal vs cancerous tissue) or among different biological mechanisms, at the gene expression level. Thus, the literature is plenty of studies, where numerous feature selection techniques are applied, in an effort to reduce the noise and dimensionality of such datasets. However, little attention is given to the stability of these signatures, in cases where the original dataset is perturbed by adding, removing or simply resampling the original observations. In this article, we are assessing the stability of a set of well characterized public cancer microarray datasets, using five popular feature selection algorithms in the field of high-throughput genomics data analysis.

HECTOR: Enabling Microarray Experiments over the Hellenic Grid Infrastructure

Biologists, medical experts, biochemical engineers and researchers working on DNA microarray experiments are increasingly turning on Grid computing with the scope of leveraging the Grid’s computing power, immense storage resources, and quality of service to the expedient processing of a wide range of datasets. In this paper we present a combined experience of grid application experts and bioinformatics scientists in deploying a pilot service enabling computationally efficient processing and analysis of data stemming from microarray experiments. This pilot service is accessible over the Hellenic portion of the EGEE grid and has been demonstrated in the scope of several public events. We highlight the process of grid application enablement, grid deployment challenges, as well as lessons learnt from a bi-annual effort to port and deploy a MATLAB DNA microarray application on a production grid. In addition to describing the parallelization of the application, we also emphasize on the development of a distributed federated database for storing and post-processing the results of the microarray experiments. Overall we believe that our experience could be proven valuable not only to microarray data scientists but also to other Grid users that intend to Grid-enable and deploy their applications.

Building in-silico pathway SBML models from heterogeneous sources

The recent revolutionary developments concerning the high throughput (-omics) measuring techniques in life sciences is expediting the way for the development of in silico models envisioning the systems biology perspective in the description of biological problems. As a result, very large open biological databases provide in silico descriptions in various formats, of biochemical pathways related to various cellular physiological aspects across the evolutionary climax. However, the lack of standardization regarding conceptual biological data representation incurs sheer limitations with respect to the functionality as well as the scientific completeness of the respective models. In this work, a software solution is presented which successfully bridges the gap towards building in-silico metabolic pathway models in Systems Biology Markup Language (SBML) format (standard SBML, CellDesigner SBML) by exploiting various XML based formats (SBML, KGML- KEGG Markup Language-, CellML - Cell Markup Language-, for pathway representation). Our solution provides methods for the biochemically correct transformation, curation and automatic simulation of the pathways, thus accomplishing the setup of fully functional in-silico models.

GRISSOM Platform: Enabling Distributed Processing and Management of Biological Data Through Fusion of Grid and Web Technologies

Transcriptomic technologies have a critical impact in the revolutionary changes that reshape biological research. Through the recruitment of novel high-throughput instrumentation and advanced computational methodologies, an unprecedented wealth of quantitative data is produced. Microarray experiments are considered high-throughput, both in terms of data volumes (data intensive) and processing complexity (computationally intensive). In this paper, we present grids for in silico systems biology and medicine (GRISSOM), a web-based application that exploits GRID infrastructures for distributed data processing and management, of DNA microarrays (cDNA, Affymetrix, Illumina) through a generic, consistent, computational analysis framework. GRISSOM performs versatile annotation and integrative analysis tasks, through the use of third-party application programming interfaces, delivered as web services. In parallel, by conforming to service-oriented architectures, it can be encapsulated in other biomedical processing workflows, with the help of workflow enacting software, like Taverna Workbench, thus rendering access to its algorithms, transparent and generic. GRISSOM aims to set a generic paradigm of efficient metamining that promotes translational research in biomedicine, through the fusion of grid and semantic web computing technologies.

GRISSOM web based grid portal: Exploiting the power of grid infrastructure for the interpretation and storage of DNA microarray experiments

DNA Microarrays have dramatically reshaped modern biological research by deriving profiles of genome-wide expression of living organisms, and producing an unprecedented wealth of quantitative data. Given this characteristic, microarray experiments are considered high-throughput both in terms of data (data intensive) and processing (computationally intensive). GRISSOM enables exploitation of GRID resources for DNA microarray distributed processing. It provides experts with a complete web-based solution for managing, searching and disseminating biological knowledge in the context of gene expression patterns on a genomic scale. The platform is developed and deployed using open source software components. Through the use of web service technologies (WSDL language) GRISSOM can be encapsulated in other biomedical processing workflows, thus rendering access to its algorithms, transparent and generic.

Using Grid Infrastructure for the Promotion of Biomedical Knowledge Mining

Transcriptomic technologies (DNA microarrays, Next generation sequencers) represent a major innovation in biomedical research contributing an unprecedented wealth of data regarding genome-wide inspection of an organism. GRISSOM web application is a microarray analysis environment, exploiting Grid technologies. In this work we present how the novel functionalities it incorporates through the use of various web services, gradually transform it to a generic paradigm for versatile biological computing, semantic mining and knowledge discovery.