[Bioperl-l] Finding all human paralogues

[Yuval Itan]

Dear Torsten,

I have solved the problem, I deeply thank you for your patience and great 
help.

All the best,

Yuval

On Friday 17 March 2006 21:13, you wrote:
> > Many thanks for your good help. I have followed your advice, and updated
> > the path to the fasta file, both in BLASTDB, BLASTDATADIR and in the
> > code. But even now it doesn't work. The directory of my fasta file
> > is: /home/Yuval/FastaDBs and the one of the Blast binaries
> > is: /home/Yuval/Applications/blast/bin
> >
> > And I keep getting the same error message:
> >
> > [blastall] WARNING: test: Unable to open
> > Homo_sapiens.NCBI354.feb.pep.fa.pin
> >
> > ------------- EXCEPTION  -------------
> > MSG: blastall call crashed: 256
> > /home/Yuval/Applications/blast/bin/blastall -pblastp  -d 
> > /home/Yuval/FastaDBs/Homo_sapiens.NCBI354.feb.pep.fa
>
> (I assume there was a space between -p and blastp)
>
> > -i  /tmp/98fwqIYOMi  -o  /tmp/vlRcHscPcH
>
> This doesn't look like a BioPerl problem anymore. It says it can't find the
> blast index files for Homo_sapiens.NCBI354.feb.pep.fa. But the "-d xxxx"
> line is trying to load them from the right place. I assume you have run
> formatdb to create the indices, and they are in /home/Yuval/FastaDBs/, and
> that there were no errors ?
>
> % cd /home/Yuval/FastaDBs/
> % formatdb -i Homo_sapiens.NCBI354.feb.pep.fa -p T -o T
> % cat formatdb.log
> % ls -lsa Homo_sapiens.NCBI354.feb.pep.fa.*
>
> and that this last "ls" lists about seven index files called
> Homo_sapiens.NCBI354.feb.pep.fa.pin, .psq, .p?? etc
> And they are READABLE by the user who is running Blast (hopefully "Yuval")?
>
> If so, can you run blast on the command line?
>
> % blastall -i SOME_PROTEIN.FA -d
> /home/Yuval/FastaDBs/Homo_sapiens.NCBI354.feb.pep -p blastp
>
> BioPerl will probably never run your blast if you can't get that working.