BioMoby services are semantic services. This means that they have service descriptions built in. This extra description layer provides extra information for connecting services together, but it also changes the way that you pass data into the services. Instead of passing in data as inputs, the data has to be first defined as a particular kind of BioMoby object type. The BioMoby objetc is the service input.
To add a BioMoby service, you must first add a BioMoby data type to the workflow. The BioMoby datatypes and services are loaded into the services panel by default and can be found in the BioMoby folder.
The data types can be found in the MOBY Objects folder. Import the datatype 'Object' into the workflow by dragging and dropping on the canvas.
Once you have added a datatype to the workflow, the 'details' tab in the workflow explorer will display what BioMoby services your object will feed into.
In the details tab, click on 'Datatype registry query' and select yes to restrict your search based on namespace. Select the namespace 'NCBI_gi' from the list.
A new window will pop up showing which services the datatype 'object' feeds into when it has a namespace of 'NCBI_gi'
Select 'getGenbankFasta' from bioinfo.icapture.ubc.ca and add it to the workflow by right-clicking.
The service will automatically be connected to the object data type.
You can chain services together in the same way. If you select 'getGenbankFasta', in the workflow diagram, the details tab in the workflow explorer will show you which services it can feed into.
Select 'getDragonBlastText'. Again, it will be automatically connected to the workflow.
Finally, find services that consume the output from 'getDragonBlastText' in the same way and add the service 'parseDragonBlastDBText'.
Your workflow should look something like this (when showing all ports)
To finish the worklow, add workflow inputs and outputs in the normal way. The finished workflow should look like this:
You can combine BioMoby services with other kinds of services by connecting to/from the appropriate BioMoby data types. To make this easier, BioMoby also provides parsers for their services so you can extract the raw data from the BioMoby data type (e.g. for getGenbankFasta, the parser can extract the ID, namespace and sequence string).
BioMoby services from BioMoby central are loaded into the services panel automatically when Taverna starts up. However, if you have your own BioMoby services, you can add them by selecting 'Import new services' in the services panel and selecting 'Biomoby services'.