Hi Sirisha We recommend all mart deployers to keep the internal names for instance in this case (pathway_db_id) in lower case. Even if some are left in mixed case, upon configuration, they are all changed to lower case.
I guess Taverna is using XMLs to get to the internal names instead of using ?type=attributes request which serves the processed names. I understand that BioMart webservice is not supplying explicit request on dataset linking, due to which Taverna is processing XMLs. Dataset Linking webservice request will soon become available - next release and guess then Taverna wont have to process XMLs to retrieve this and other metadata info. Sirisha, hope you can do the necessary modifications for now to get your workflow going. David, thanks very much for spotting this. Cheers Syed On Thu, 2008-11-20 at 13:56 +0000, David Withers wrote: > config -- ====================================== Syed Haider. EMBL-European Bioinformatics Institute Wellcome Trust Genome Campus, Hinxton, Cambridge CB10 1SD, UK. ======================================
