From Documentation
Revision as of 10:31, 26 March 2012 by Isaac (Talk | contribs)

Jump to: navigation, search
BLAST
Description: To find regions of local similarity between sequences.
SHARCNET Package information: see BLAST software page in web portal
Full list of SHARCNET supported software



NCBI BLAST is installed under _/opt/sharcnet/blast/current_ on all of SHARCNETs XC clusters as well as various other clusters specified in the Availability Table.

The current directory for BLAST contains four sub-directories named _bin_, _data_, _doc_ and _lib_, where the data directory contains BLOSUM and PAM matrices. This data directory can also be shared with the mpiBLAST package as discussed in the mpiBLAST FAQ. The _bin_ directory contains all the blast binaries, such as @blastall@. This directory is included in your default search path.

Since blast is a serial application jobs should ideally be only run on whale cluster. For example serial jobs can be submitted as follows:

sqsub -r 5d -o ofile%J blastall -p blastn -i ech_query.fas -d yeast.nt -o blastn_test.typ
sqsub -r 5d -o ofile%J blastall -p blastx -i ech_query.fas -d yeast.aa -o blastx_test.typ
sqsub -r 5d -o ofile%J blastall -p tblastn -i test3.aa -d yeast.nt -o tblastn_test.typ
sqsub -r 5d -o ofile%J blastall -p blastp -i test3.aa -d yeast.aa -o blastp_test.typ
sqsub -r 5d -o ofile%J blastall -p tblastx -i ech_query.fas -d yeast.nt -o tblastx_test.typ

Before running jobs, create a _~/.ncbirc_ file in your home account containing something like:

[NCBI] 
Data=/opt/sharcnet/blast/current/data
[BLAST]
BLASTDB=/work/username/your/db/directory
BLASTMAT=/opt/sharcnet/blast/current/data

Features available in the currently installed version(s) are shown "here"[1].