	Sample BLIMPS files (Block Searcher)


S. Henikoff & J.G. Henikoff, "Protein family classification based on 
searching a database of blocks." Genomics 19:97-107 (1994). 
http://blocks.fhcrc.org/blocks_search.html


Three C-shell scripts here explain how to execute BLIMPS two ways
followed by three post-processing programs:

Search a sequences vs a blocks database (Block Searcher):
1. blimps.blkprob.csh: executes blimps followed by blkprob interactively.
Just accept all the defaults. blkprob is the current post-processing
program and gives combined E-values for each hit.

2. blimps.blksort.csh: executes blimps followed by blksort,
the original post-processing program which gives rank statistics.
Just accept all the defaults.

Search a set of blocks vs a sequence database:
3. multimat.csh: This script is not interactive, just execute & examine
the output files. multimat is not easy to use as it requires precise
file names, which are explained within the script. You may find it easier
to use Tim Bailey's MAST program to search blocks vs a sequence database.
The program $BLIMPS_DIR/bin/blk2pssm will prepare the blocks for input to
mast (blk2pssm <input_file_of_blocks> <output_file> M). MAST may be
obtained at http://www.sdsc.edu/MEME .

To test implementation, run the *.csh files, accepting all the default
answers, and compare the output files with those in distribution/ .

-------------------------------------------------------------------------

Two perl scripts here demonstrate how to run bulk searches:

batch-search.pl
       Sample script to do bulk Block Searches locally
       Argument is <file of query sequences in FASTA format>

www-search.pl
       Sample script to do bulk Block Searches on the web server
       Replaces Email Block Searcher
       Arguments are <file of query sequences in FASTA format> <return email>
       Requires libwww-perl         (http://www.linpro.no/lwp/)

