How to Run WU-BLAST
This page describes the requirements for using WU-BLAST to align one or more query
sequences with a database of target sequences.
WU-BLAST is installed on Beagle and Lewis.
To test if the WU-BLAST tools are in your path type "wu-blastall" (without the quotes) at the command line.
Options available for WU-BLAST should be listed.
More options for a given type of WU-BLAST can be found by entering the type of blast (e.g., blastn).
The path to WUBLAST on Beagle and Lewis is /dfs/bio/wublast.
To run WU-BLAST efficiently the command must be put into a queue.
On Beagle this is done with
PBS (Portable Batch System)
and on Lewis with LSF (Load
Sharing Facility).
The script enqueued by PBS or LSF should contain at least the following command and arguments:
blast-command /database_path/database query-filewhere blast-command is either blastn, blastp, blastx, tblastn, or tblastx, and should be prefaced by /dfs/bio/wublast if it is not in your path already. The path to WU-BLAST-formatted genbank is /dfs/databases/genbank and all (gball) or any division (e.g., gbest) of genbank may be used as a target. The query-file contains all query sequences in fasta format. For a more complete description of the WU-BLAST options go here.
To create your own WU-BLAST-formatted database use the xdformat command. Entering the command alone will return the directions to do this.
Slides and notes from a workshop on BLAST and other types of sequence similarity searching may be found at the UMBC's Education site.