A series of short talks on the development of ontologies

To wet the appetite for the upcoming UK Ontology Network meeting on the 14th of April, ICOS presents a series of short ontology related talks:

Workflows for developing biological ontologies using a document-centric approach

Tools used by ontology developers are less understood by the domain specialists who use different formalisms to represent and manipulate their data. Existing tools, such as WebProtege and Populous require leaving the tool environment to move into ontology specific environment. We are investigating a document-centric workflow centred around the use of English and standard office software which should ease the interaction between a domain specialist and an ontologist. In this proposed worklflow, ontologists will generate a computational representation which can be transformed into documents that biologists can then read and correct. An arbitrary and frequent transformation of documents will be performed using Tawny-OWL, as well as maintaining the original files as part of the ontology source code. Currently we are testing these workflows using as a source the tolergenic dendritic cell catalogue in a spreadsheet format.

DReNIn_O: A high-level ontology for drug repositioning

Drug development is both increasing in cost while decreasing in productivity. There is a general acceptance that the current paradigm of research and development needs to change. One complementary approach is that of drug repositioning which focuses on the identification of novel uses for existing drugs. Many marketed examples of repositioned drugs have been identified through serendipitous or rational observations, highlighting the need for more systematic methodologies. Systems approaches have the potential to enable the development of novel methods to understand the action of theraputic compounds, but require an integrative approach to biological data. Here we present DReNIn_O, an application ontology for drug repositioning with the aim of making the relevant integration task easier. DReNIn_O represents data describing drugs in relation to their affect on targets and diseases. Developed to aid the integration and subsequent mining of drug repositioning networks, DReNIn_O semantically describes relationships between 25 data types relevant to drug repositioning. These types include: Disease (with child terms Rare_Disease and Common_Disease); Drug_Molecule (with child terms including Small_Molecule); Biological_Molecule (with child terms including Protein and Gene); and Annotation (with child terms including Clinical_Trial). An integrated RDF dataset that makes use of DReNIn_O is also presented.

The Synthetic Biology Open Language

Synthetic biology has great potential to develop novel biological systems. As in many engineering disciplines, synthetic biologists also use techniques such as standardisation, modularity and abstraction, in order to develop predictable applications and to scale up designs. However, engineering biological systems is not a trivial task. These systems can be complex and they can have very large design spaces. Identified possible solutions are verified through design-build-test cycles, which may involve several tasks and require inputs from different researchers. Computational tools are especially useful to identify these solutions and to facilitate the execution of these tasks. However, tools often focus upon particular aspects of design processes and work in isolated manner. Data standards are needed to unambiguously exchange information between these tools in order to facilitate the execution of complex workflows, and to increase the reproducibility of designs across different labs.

The Synthetic Biology Open Language (SBOL) is a data standard that has been specifically developed to exchange biological system designs. Using this language, biological systems can be defined with individual design components, which can represent DNA, RNA, proteins or metabolites. SBOL facilitates the reuse of these components and existing complex designs. Desired roles of each component and possible interactions can also be specified. Moreover, SBOL allows incorporating application specific information. SBOL utilises existing Semantic Web resources to represent, store and retrieve data. Designs are serialized using the RDF/XML format. Although SBOL defines its own terms, it reuses existing controlled vocabularies and ontologies, where possible. For example, terms from the BioPAX ontology are used to indicate types of design components and Sequence Ontology terms are used to indicate roles of DNA-based components. Application specific data can either be embedded as RDF triples within SBOL entities or defined as RDF resources at the top level. Although, such custom data may not be understood directly by SBOL libraries, they are retained during read and write processes. SBOL is aimed directly for tool developers; however, through the development of SBOL compliant tools wet-lab biologists can take the advantage of many useful tools in order to create and refine their designs. The development of SBOL is carried out openly by the SBOL Developers group. This development is coordinated by elected editors and an SBOL chair, and a steering committee. Currently, SBOL Developers include more than 120 members from 52 organisations in 15 countries.