Experiences with e-Science workflow specification and enactment in bioinformatics

Workflow techniques form an important part of in-silico experimentation within the bioinformatics domain and potentially allow the eScientist to describe and enact their experimental processes in a structured, repeatable and verifiable way. Bioinformaticians routinely use Web-based resources within their in-silico experiments. However, the use of current web service orchestration techniques is problematic, and represents a significant barrier to take-up by the bioinformatics community, due to the rapidly evolving and competing standards, a lack of freely available tools, limited support for interaction with stateful services, and inappropriate levels of abstraction for the bioinformatics domain. As a result, the EPSRC funded Grid[11] project has, in collaboration with the European Bioinformatics Institute and the Human Genome Mapping Project, developed a graphical toolset and workflow enactor which uses its own high level representation of a process flow, including specification of processing units, data transfers and execution constraints.