By the way to follow-up to an answer, please use comments. Don't create another answer. You will want to move your "answer" to a comment on the answer to which it relates.
Your error from PSIPRED looks like the one discussed in "Trying To Run Psipred But Failing To Use Blastpgp", and from your description it sounds like the steps required to create the filtered database used by PSIPRED to perform the PSI-BLAST searches were omitted. From a quick check in the current PSIPRED release (pgppsipred3.5.tar.gz), it appears that the 'runpsipred' script is configured to use the UniRef90 based 'uniref90filt' database by default, instead of the NCBI nr based 'nrfilt' database described in the README file. So I take it that you have modified the 'runpsipred' script to use the database you have, and have ensured that it points to the correct location, or that the location is set in the $BLASTDB environment variable.
If you still have problems you will need to know what the actual PSI-BLAST command being run is, so you can debug it in isolation (and see any error messages from PSI-BLAST). You can do this by careful examination of the 'psipred' script, or by adding the '-x' option to the shebang line in the script so all the commands are reported by the shell, for example:
$ ./runpsipred test_seq.fasta
set dbname = uniref90filt
set ncbidir = /usr/bin
set execdir = ./bin
set datadir = ./data
set basename = test_seq
set rootname = test_seq
set hostid = `hostid`
hostid
set tmproot = psitmp3300007f0101
cp -f test_seq.fasta psitmp3300007f0101.fasta
echo Running PSI-BLAST with sequence test_seq.fasta ...
Running PSI-BLAST with sequence test_seq.fasta ...
/usr/bin/blastpgp -b 0 -j 3 -h 0.001 -d uniref90filt -i psitmp3300007f0101.fasta -C psitmp3300007f0101.chk
if ( 1 != 0 ) then
tail psitmp3300007f0101.blast
echo FATAL: Error whilst running blastpgp - script terminated!
FATAL: Error whilst running blastpgp - script terminated!
exit 0
You can now run the PSI-BLAST command manually to debug the issue. In this particular case the database is not found and 'blastpgp' exits with a 130 exist status.
I have same problem with psspred
You mean if I use uniref90filt the problem will solve?
No.
By the way to follow-up to an answer, please use comments. Don't create another answer. You will want to move your "answer" to a comment on the answer to which it relates.