How-to BLAST
This page will tell you how to run NCBI's Basic Local Alignment Search Tool (BLAST) using a command line prompt. NCBI also has a web-based BLAST that you can run here.
1. Download the BLAST+ tool kit for your operating system.
2. Download the necessary database.
-For the Sawyer et al. (2011) paper, we wanted the non-redundant (nr) database of all known protein sequences—nr.00.tar.gz, etc. For the directory of databases, see this link.
3. Obtain query file in FastA format.
Ex.
>Name
SEQUENCEEITHERPROTEINORDNA
4. Choose your parameters.
The BLAST command
You code a BLAST command by giving parameters designated by a single letter with a dash in front. Generally, lowercase letters refer to the input and uppercase letters refer to the output. Our BLAST call:
-designates where the computer can find the program and which version of the program to use (-p, blastp for protein BLAST)
-the name of the input (-i)
-the database to BLAST against (-d, name includes where to find that database)
-type of statistics to use (-C, computational)
-format of output (-m)
-name of output (-o)
We add on other information that is specific to our computer, such as how many processors to use and courtesy settings if other people need to run a program.
An example call:
blastall (call blastall program) –p (type of input, -p means protein) blastp (sub-program, run protein blast) –b 100 (maximum 100 hits, default is 250) –i input.fa (input file) –d nr (databasefile) –m 8 (output format option 8) –o output.txt (output file) –e 0.01 (quality control level e=0.01)
Our call for the Sawyer et al. (2011) paper:
/localapps/blast-2.2.18/bin/blastall -p blastp -i input.fa -d data/gb/nr -C 2 -m 8 -o output.blast