Standalone blast

[jomaco]

Hi

I have downloaded standalone blast and formatted a nucleotide database. I wish to use tblastn to compare Arabidopsis proteins against it. I made a db folder in the folder where blast is saved and changed to that directory then ran the following command:

tblastn -query tair1prot -db petunia -evalue 0.01

However I receive an error:

BLAST Database error: No alias or index file found for the nucleotide database [petunia] in search path [C:\Program Files\NCBI\blast-2.2.25+\db;C:\Program Files\NCBI\blast-2.2.25+\db;]

As you can see I think the program is taking the file path of the current directory and then adding the directory where databases are stored on to the end. This would mean the db directory and the index file are not found. If I change directory to C:\ for instance then the errror would be C:\;C:\Program Files\NCBI\blast-2.2.25+\db;

How can I stop this from happening so the database can be located correctly? What do you think I might be doing wrong?

Thanks

Jon

[maubp]

The simplest answers are either tell BLAST where you files are (instead of -db petunia use something like -db "C:\my files\petunia"), or put your files in the BLAST database folder where BLAST automatically looks, C:\Program Files\NCBI\blast-2.2.25+\db

[bioBob]

did you format your database? I get the same error when I forget to run formatdb

[jomaco]

Hi

Thanks for your replies.

I was using makeblastdb which I think formats the db so I think my problem was a permissions issue. I ran tblastn from a folder in my documents and the output file was created there (it wouldn't work when running from within program files).

I've had success running one or two proteins against the nucleotide database. However, now I've started the program with the entire TAIR protein database it is understandably taking a long time to run. My nucleotide database is about 1.3MB in length and I believe there are ~300000 proteins. How long can I expect this to take on a normal pc (i7 processor/4GB RAM) or should I really be thinking about running it on a supercomputer?

Thanks

Jon