Update: Downloading Preformatted BLAST databases

Databases for use with BLAST have traditionally been generated from text files of sequences in FASTA format using the BLAST utility program ‘formatdb’. As the sequence databases have grown, standard BLAST databases in FASTA format have become increasingly unwieldy to download and store locally. To help alleviate the burden of maintaining these large FASTA files locally, NCBI provides preformatted BLAST databases at:

ftp.ncbi.nih.gov/blast/db

Although these preformatted databases have been available for a few years, many people continue to download the FASTA formatted databases. This is quickly becoming a Herculean feat as the largest of these FASTA files now exceeds 40 Gigabytes. Using preformatted files has several advantages — shorter download times, no need for running formatdb, and inclusion of taxonomy information for database entries. If needed, the entire FASTA formatted version of the database or any subset of sequences can quickly be generated from the formatted database using the BLAST utility fastacmd (See BLAST Lab, NCBI News, Summer 2003). For instance,

fastacmd -d nr -D1

extracts the entire preformatted nr database in FASTA format.

Larger formatted databases such as nt, est and wgs are broken into smaller one-Gigabyte volumes. A PERL script is available to simplify this task of keeping preformatted BLAST database current.