Publications and Presentations

Semantic inferencing and querying across large-scale RDF triple stores is notoriously slow. Our objective is to expedite this process by employing Google’s MapReduce framework to implement scale-out distributed querying and reasoning. This approach requires RDF graphs to be decomposed into smaller units that are distributed across computational nodes. RDF Molecules appear to offer an ideal approach – providing an intermediate level of granularity between RDF graphs and triples. However, the original RDF molecule definition has inherent limitations that will adversely affect performance. In this paper, we propose a number of extensions to RDF molecules (hierarchy and ordering) to overcome these limitations. We then present some implementation details for our MapReduce- based RDF molecule store. Finally we evaluate the benefits of our approach in the context of the BioMANTA project – an application that requires integration and querying across large-scale protein-protein interaction datasets.

Semantic inferencing and querying across large scale RDF triple stores is notoriously slow. Our objective is to expedite this process by employing Google’s MapReduce framework to implement scale-out distributed querying and reasoning. This approach requires RDF graphs to be decomposed into smaller units that are distributed across computational nodes. RDF Molecules appear to offer an ideal approach – providing an intermediate level of granularity between RDF graphs and triples. However, the original RDF molecule definition has inherent limitations that will adversely affect performance. In this paper, we propose a number of extensions to RDF molecules (hierarchy and ordering) to overcome these limitations. We then present implementation details for our MapReduce-based RDF molecule store describing: (a) graph decomposition into molecules; (b) SPARQL querying across molecules; and (c) molecule merging to retrieve the search results. Finally we evaluate the benefits of our approach in the context of the BioMANTA project – an application that requires integration and querying across large-scale protein-protein interaction datasets. The results of performance evaluations based on this case study are presented and discussed.

The computational analysis of protein-protein interaction and biomolecular pathway data paves the way to efficient in silico drug discovery and therapeutic target identification. However, relevant data sources are currently distributed across a wide range of disparate, large-scale, publicly-available databases and repositories and are described using a wide range of taxonomies and ontologies. Sophisticated integration, manipulation, processing and analysis of these datasets are required in order to reveal previously undiscovered interactions and pathways that will lead to the discovery of new drugs. The BioMANTA project focuses on utilizing Semantic Web technologies together with a scale-out architecture to tackle the above challenges and to provide efficient analysis, querying, and reasoning about protein-protein interaction data. This paper describes the initial results of the BioMANTA project. The fully-developed system will allow knowledge representation and processing that are not currently available in typical scale-out or Semantic Web databases. We present the design of the architecture, basic ontology and some implementation details that aim to provide efficient, scalable RDF storage and inferencing. The results of initial performance evaluation are also provided.

Protein-protein interaction (PPI) and biomolecular pathway data hold tremendous potential for drug discovery and development. However, relevant data sources are currently distributed across a wide range of disparate, large-scale, publicly- available databases, web sites and repositories and are described using a wide range of taxonomies and ontologies. Sophisticated integration, manipulation, processing and analysis of these data sets are required in order to reveal previously undiscovered interactions and pathways that will lead to the discovery of new drugs. The Semantic Web has been investigated as a solution to this problem by a number of projects that use RDF and OWL to integrate, represent and analyze protein interaction data. However, existing applications have suffered from certain limitations that hinder their usefulness. In this paper, we describe work being undertaken within the BioMANTA project that aims to identify and overcome the limitations associated with the application of Semantic Web technologies to protein interaction network analysis. In particular we describe the BioMANTA OWL ontology that has been designed to enable multiple data sources to be integrated within a single RDF triple store through a common PPI model. The primary aim of the BioMANTA ontology is to provide a practical means of facilitating the integration of semantically disparate data sets – it does not aim to provide a precise biological, chemical or physical model of how proteins interact. We also describe how this ontology was developed through the refinement, harmonization and extension of existing ontologies. Finally, we describe the mapping, integration and querying of a range of protein-protein interaction data sets, based on the ontology.

M.Davis, A.Newman, I.Khan, J.Hunter, M.A. Ragan, "Integrating Hierarchical Controlled Vocabularies with OWL Ontology: A Case Study from the Domain of Molecular Interactions", accepted to 6th Asia Pacific Bioinformatics Conference (APBC08), Kyoto, Jan 14-17, 2008.