Combining the Power of Taverna and caGrid: Scientific Workflows that Enable Web-Scale Collaboration

Service-oriented architecture represents a promising approach to integrating data and software across different institutional and disciplinary sources, thus facilitating Web-scale collaboration while avoiding the need to convert different data and software to common formats. The US National Cancer Institute´s Biomedical Information Grid program seeks to create both a service-oriented infrastructure (caGrid) and a suite of data and analytic services. Workflow tools in caGrid facilitate both the use and creation of services by accelerating service discovery, composition, and orchestration tasks. The authors present caGrid´s workflow requirements and explain how they met these requirements by adopting and extending the Taverna system.