Exploring small molecules and their drug-like properties using Semantic Web technologies

Joo Young Choi (2010). Exploring small molecules and their drug-like properties using Semantic Web technologies MPhil Thesis, Institute for Molecular Bioscience, The University of Queensland.

Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
s41276198_mphil_finalthesis.pdf final thesis application/pdf 10.63MB 0
Author Joo Young Choi
Thesis Title Exploring small molecules and their drug-like properties using Semantic Web technologies
School, Centre or Institute Institute for Molecular Bioscience
Institution The University of Queensland
Publication date 2010-02
Thesis type MPhil Thesis
Supervisor Mark A. Ragan
Melissa J. Davis
Total pages 121
Total colour pages 10
Total black and white pages 111
Subjects 06 Biological Sciences
Abstract/Summary There is currently great interest in the application of computational methods to chemical data such as small molecules because they play important roles in biology, chemistry and the pharmaceutical industry. Not only information about chemical structures and their measured properties, but also data about biological assays and targets have increasingly come into demand, and the number of online resources containing such chemical data has grown in parallel with the rapid development of the World Wide Web. However, there are many restrictions in the presentation of or access to such data, and the resources make varying use of standards and conventions. These problems arise from the lack of open data or social collaboration, and the lack of persistent standardisation for knowledge representation and biological terminology on the Web. The Semantic Web reflects a new vision for the World Wide Web in which online resources are presented in a way that enables computers to interact more intelligently with data they contain. This vision has potential to improve the representation of chemical data and to enable the application of machine inference and other computational methods to facilitate knowledge discovery and improve the utility of existing data collections such as existing stores of chemical data. The use of Semantic Web technologies may enable the expression of chemical and small-molecule data in formats that remove obstacles, and provide sufficiently high-quality metadata for automated the application of analysis methods. In this research I introduce a new Small Molecule Ontology (SMO) that represents concepts and includes structural information and unique identifiers such as IUPAC names for biologically relevant properties of small molecules, and of their interactions with biomolecules such as proteins, using the Semantic Web technologies RDF, RDFS and OWL. Instance datasets have been generated for the SMO from heterogeneous public data sources integrating chemical and biologically relevant data by conversion into RDF triples held in an RDF triple-store. Through the use of pre-defined competency questions implemented as SPARQL queries in my research, I have evaluated SMO and have demonstrated that small-molecule and biomolecular data sources are effectively and flexibly represented, and that useful knowledge can be extracted and inferred. To demonstrate the potential applicability of SMO in support of computational reasoning in drug discovery, I applied criteria for evaluating the drug-likeness of small molecules, including Lipinski’s Rule of Five and similar rule sets, to the SMO, and then implemented these rule sets as SPARQL queries and extracted drug-like molecules from the instance dataset of the SMO. I show that Lipinski’s rule has higher accuracy rather than others, but does not retrieve all known drug-like molecules. Nonetheless its high success rate indicates that the rule can be a valuable tool, and this work demonstrates the capacity of Semantic Web technologies to manipulate chemical data with potential application in pharmacogenomics research.
Keyword Small molecules
Semantic Web
Additional Notes 14-16, 31, 57-61, 86

Citation counts: Google Scholar Search Google Scholar
Created: Wed, 08 Jun 2011, 15:50:26 EST by Ms Joo Young Choi on behalf of Library - Information Access Service