Provide methods for doing micro array analysis using Taverna, for instance by using methods from Expression Profiler and running R scripts using Bioconductor.
The micro array community also wants to be able to run Taverna as a Grid service or Web service, to be able to monitor and control progress programmatically.
Overall Goals
Micro array analysis workflows could possibly take a long time to execute, and therefore some way of running a workflow detached is important, preferably as a web/grid service.
R is used to a large extent in the community, so a R processor is important. However, R is usually only applied to run BioConductor services, and so what is really needed is to expose bio conductor services to Taverna
The Expression Profiler provides many micro array analysis services. These services should be usable from Taverna, but this might be easiest to achieve if the EP people make a WSDL based web service of the EP.
Micro array data can be huge, so Taverna must be able to handle this, preferably by some referencing or data caching scheme.
When testing several T-test values, it would be useful to do a parallel iteration over the different values. Some way of "submitting" the job to a cluster of web services instead of a single web service.