I gave John, Tim and Ron a brief demonstration of Taverna. Proxy settings that I used are found in runme3.bat, which is attached to this page. The demonstration was on a 128mb win98 pc. I demonstrated workflow SRSQueryUsingUSAAndFormat and ClustalwLocalConcat which can be found here. Tim indicated that he thought that these workflows would be useful for their work, but that the clustalw example could be made more useful. He indicated that rather than taking two accession numbers as input, looking up their sequences and then aligning these sequences, it would be more useful to take multiple accession numbers, look up all the sequences for all of them and align them against each other. He indicated that another useful workflow would take as input the name of a protein, look up all the proteins whose name matched this, then look up the nucleotide sequences from which these proteins were translated and align these against each other.

After the interview, I wrote AccNumListToClustalw which does a multiple alignment given a list of accession numbers, and Peter Li wrote ProteinNameToCrossReferencedAccessionNumbers to look up nucleotide sequences for a given protein name. They can be found here.

John and Tim indicated they were interested in working with Taverna, so I left them an installation on their computer. They have suggested that they provide me with a list of common problems which they come across that could be solved using workflows.