On Saturday, May 7, 2011, <drhoads@uark.edu> wrote:
Thanks Peter, I debated which to post to. This is just using the Megablast tool in Galaxy. My IT people are not Biologists. They had this working fine and some tweak has screwed up the GI column in the results file. It should be the same NR and HTGS datafiles that we had before. So you recommend that I should just have them reinstall the databases and make sure to use the correct switches?
You don't know what "tweak" broke things? You've got at least three factors, (1) Galaxy and the wrapper, but I don't see how an update of these would have affected the IDs. (2) The BLAST command line tool (blast all in this case), perhaps it was updated and this has altered the output? (3) The BLAST database(s) were updated (or rebuilt if you have local databases, in which case it could have been done with a newer version of the formatdb tool). If you are using NR does you have it regularly updated from the NCBI FTP site? Hopefully your admin team will have some sort of log of system updates that might help work out what happened. Peter