Hi John, I've CC'd the mailing list again (assuming you omitted it by mistake - its easily done). On Tue, Nov 9, 2010 at 8:43 PM, John David Osborne wrote:
I've written a BLAST+ wrapper which is now in galaxy-central, and will eventually be in galaxy-dist and thus potentially on the public Galaxy server (assuming it won't tax it too much). This doesn't (yet) offer the option to run the BLAST remotely at the NCBI (the wrapper could do this in theory). However, running a BLAST against the NCBI databases will cause difficulties for reproducibility (since we have no control over the databases, and the NCBI make regular updates).
Understood, I assume then most people get their sequence data and upload before starting a workload... That wrapper does sound handy though! It just seems odd to me that the Galaxy portal doesn't have a generic blast tool and standard (nt, nr, etc...) to run against.
Well, I'm sure they will consider it in future. But if I were in their shoes, I'd be a little nervous about the computational load it might impose.
Would you be running your own Galaxy server?
That's the current plan. We are also looking at Taverna, but it doesn't seem like I can find much in the way (actually none) of biology publications that use it as part of their workflow. I haven't looked that hard though.
Within our institute we intend to offer BLAST running on our departmental server (in Galaxy) for Biologists to run large searches (e.g. multiple query sequences like a set of contigs) against both NCBI databases like NR and also in house ones (e.g. unpublished genomes), and as part of workflows (e.g. upload a FASTA file, blast against organism X, divide the FASTA file into those with good matches and those without). This is drifting into a discussion more suited for the galaxy dev mailing list though ;) Peter