# Parse_PSI

This is a basically a fix for blast versions greater than 2.0.3

This little script allows parsing of psiblast report files
to extract the embedded sequence alignment, and write out
in different formats. The fasta format alignment can then
be viewed in Jalview et. al.

change_header will convert the perl headers if you need it

parse_psi -help  : help
parse_psi -degap : remove all gaps in alignment and print as
	           a fasta format file
parse_psi -ungap : remove gaps in query sequence, and those
                   below the gap in the query sequence, print
                   as fasta format
parse_psi -gap   : print alignment with gaps in fasta format
parse_psi -freq  : print frequency profile needed for Jnet 

Usage:

parse_psi -freq file.blast > file.freq


BLAST Notes for use with Jnet:
------------------------------
For success with parse_psi, run psiblast with following parameters:
./blastpgp -Q test.pssm -m6 -d ./pdb -b20000 -a5 -j3 -i test.seq > ! test.blast

            ^            ^        ^      ^    ^   ^    
        pssm file        1        2      3    4   5   

1 - *needed* this prints the alignments in the right format for Jnet
2 - database
3 - needs to be a big number so that blast prints all the available 
    sequences in the alignment in the right format
4 - number of processors
5 - iterations

getpssm now works with the -Q option in psiblast to get the
pssm file needed for Jnet:

getpssm test.pssm > test.pssm.jnet


