I've now understand how to integrate those tools in Taverna, a
workflow engine (http://taverna.sourceforge.net) (see the screenshot
in attachement). What is important to understand here, is that people
using the WebService & Tarverna don't need a local copy of operon.
The test I wrote takes "rs25" as an input, calls Operon to get the
coordinate of the snp, extends the coordinates to 100bp , fetches the
set of SNPs in this region and saves the result in a local file.
<?xml version="1.0"?>
<ns2:getSNPByPositionResponse xmlns:ns2="http://ws.server.operon.cephb.fr/">
<SnpList>
<acn>rs12699208</acn>
<chromosome>Chr7</chromosome>
<name>rs12699208</name>
<position>11549694</position>
<sequence>TATTCCTTAAGGTTAGATCTAAAATTCTATAG(...)
</SnpList>
<SnpList>
<acn>rs27</acn>
<chromosome>Chr7</chromosome>
<name>rs27</name>
<position>11549750</position>
<sequence>ACTAGATTTGTGTCACATATGCAGATAATG(...)
</SnpList>
(... #### ....)
</ns2:getSNPByPositionResponse>
I've described how to deploy the WS and to create this workflow on my
blog: http://tinyurl.com/r7eo62
This workflow was also uploaded and shared on myexperiment.org
http://www.myexperiment.org/workflows/757
Pierre