Automating Experiments Using Semantic Data on a Bioinformatics Grid

  • Authors:

  • Chris Wroe;Carole Goble;Mark Greenwood;Phillip Lord;Simon Miles;Juri Papay;Terry Payne;Luc Moreau

  • Affiliations:

  • University of Manchester;University of Manchester;University of Manchester;University of Manchester;University of Southampton;University of Southampton;University of Southampton;University of Southampton

  • Venue:
  • IEEE Intelligent Systems
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

myGrid assists bioinformaticians in designing and executing in silico experiments using the Grid's resources. In myGrid, much of this experimental design has been encoded as workflows. Workflows must be represented at tiered levels of detail to ensure that they can respond to changes in service availability, be customized to services in different locations, and be shared with others to varying degrees. The authors have developed workflow templates in which classes of services are composed, and a resolution mechanism by which these classes are instantiated. The specification of service classes and their resolution depends on seven kinds of service metadata. Functionally equivalent services vary widely in implementation. The authors describe workflow harmonization in which the workflow is modified to accommodate variations between substituted services. Finally, they examine the role of scientist and automated process in resolution and harmonization and discuss scope for further automation.