In addition to providing BLAST sequence alignment services on the web, NCBI also makes sequence alignment utilities available for download through FTP. This allows BLAST searches to be performed on local platforms against databases either downloaded from NCBI or created locally. These utilities are run through command (DOS-like) windows and accept text-based commands. There is no graphic user interface.

The following tutorial discusses the steps needed to install NCBI C++ based BLAST programs (blast+) and a sample NCBI database on PCs running Windows Operating Systems. The installation of soon to be deprecated C-based BLAST package (legacy blast) will be discussed only briefly at the end.

The blast+ software package is available as two self-extracting archives. The ncbi-blast-#.#.#+-win32.exe is compatible with PCs running Windows XP, 32-bit Vista or Windows 7. The other archive, ncbi-blast-#.#.#+-win64.exe, is for PCs with 64-bit chipsets running Windows XP 2003, 64-bit Vista or Windows 7. In both cases, "#.#.#" denotes the current version number of the package. Archives with the same base name are equivalent.

Note: the archive with the ".tar.gz" file extension does not have the installer function. The discussion below focuses on archives with an ".exe" extension.

Steps

Steps to download the package are described below.

• Point a browser to this ftp directory:
  ftp://ftp.ncbi.nlm.nih.gov/blast/executables/LATEST/

• Right click on a desired archive and select "Save link as…" from the popup menu

• In the prompt, switch to a desired directory (folder) and click the "Save" button to save the archive to a desired location on the local disk

The above steps, as sample screenshots for the "ncbi-blast-2.2.23-win32.exe" archive, are given in Figure 2, where the first two steps are demonstrated by the top panel and the last step is demonstrated by the bottom panel.

Figure

Figure 2a. Downloading of a blast+ package from NCBI through a web browser: Log on to ftp://ftp.ncbi.nlm.nih.gov/blast/executables/LATEST/ and select "Save link as ..." after right-clicking on "ncbi-blast-2.2.23+-win32.exe".

Examples

Example downloading steps are depicted by screen shots in Figure 2a and 2b.

Figure

Figure 2b. Downloading of a blast+ package from NCBI through a web browser: Change the location in the subsequence Windows prompt to "C:" before saving the archive to a desired location.

The blast+ archive downloaded above contains a built-in installer. Accepting the license agreement after double-clicking, the installer will prompt for an installation directory. The default installation directory in this test case is "C:\blast-2.2.23+". Clicking the "Install" button, the installer will create a "doc" subdirectory containing a comprehensive user manual in pdf format, an "uninstaller" for future removal of the installation, and a "bin" subdirectory where the following BLAST programs and accessory utilities are kept.

blastdbcheck.exe      blastdbcmd.exe            blastdb_aliastool.exe
blastn.exe            blastp.exe                blastx.exe
blast_formatter.exe   convert2blastmask.exe     dustmasker.exe
legacy_blast.pl       makeblastdb.exe           makembindex.exe
psiblast.exe          rpsblast.exe              rpstblastn.exe
segmasker.exe         tblastn.exe tblastx.exe   update_blastdb.pl
windowmasker.exe

For functions of individual program or utility, refer to Table x in Standalone BLAST setup for Unix.

Test BLAST database

In addition to BLAST programs and accessory utilities, target database are also a key component of BLAST searches. The common set of pre-formatted NCBI BLAST databases is available as compressed archives from NCBI ftp site, or databases can be prepared de novo from custom FASTA sequences locally using the makeblastdb utility. For more effective database management, a "db" subdirectory should be created under the "C:\blast-2.2.23+\".

Steps for downloading preformatted BLAST databases from NCBI are:

• point a browser to ftp://ftp.ncbi.nlm.nih.gov/blast/db/

• right-click on a desired file (refseq_rna.tar.gz in this example case)

• select "Save link as …" from the popup menu

• prompted, use the "Save in" to change the directory to "C:\blast-2.2.23+\db\"

Similar procedures in Figure 2 can be used to download the BLAST databases.

This downloaded database is blast-ready after inflation and extraction with a decompression utility. Figure 3 below shows an example inflation/extraction procedure using WinZip.

Figure

Figure 3. Exacting the downloaded refseq_rna.tar.gz archive using WinZip.

A utility included in the blast+ package, update_blastdb.pl, is designed for regular downloading of preformatted BLAST databases from NCBI. It requires the installation of the Perl package and execution from the command prompt under the "C:\blast-2.2.23\db\" directory. The base command is:

perl update_blastdb.pl --passive base_database_name

where "base_database_name" is the name of the target database, without the "##.tar.gz" extension.

For smooth execution of blast+, the PC must be configured to recognize BLAST programs under the "C:\blast-2.2.23+\bin\" directory. To do this, a user environment variable named Path needs to be created with "C:\blast-2.2.23\bin\" as its value. The blast+ installer automatically creates a BLASTDB environment variable, which points to the "C:\blast-2.2.23+\" directory. For this setup with a separate "db" directory created for BLAST databases, the value of BLASTDB environment variable must be modified to "C:\blast-2.2.3+\db\".

Example Screen Shots

Screen shots of these steps are given in Figure 4a, 4b, 4c, and 4d.

Figure

Figure 4a. Configure standalone BLAST using Windows' environment variables:
The initial System popup displaying the system information.

Figure

Figure 4b. Configure standalone BLAST using Windows' environment variables:
Clicking the "Advanced" tab to see the "Environment Variables" button, which provides access to the Environment Variables settings.

Figure

Figure 4c. Configure standalone BLAST using Windows' environment variables: Clicking "Environment Variables" button on 4b leads to this popup, where existing environment variables can be accessed or new ones can be created using the "New" button. In this (more...)

Figure

Figure 4d. Create new user variable using the "New" button: Clicking the "New" button in Figure 4c leads to a popup similar to this where new variable's name and path can be specified. Here a new Path variable is being specified with a value of "C:\blast-2.2.23+\bin\". (more...)

Standalone blast+ programs do NOT have a graphical user interface (GUI) and must be executed from a command prompt window (CMD). This window can be opened by clicking on "Start Program Accessories Command Prompt" or by clicking "Start Run," typing "cmd" (minus quotes) and pressing enter. These processes are shown in the following screenshots (Figures 5.1a and 5.1b).

Figure

Figure 5.1a Open a command prompt under Windows (XP): using the "Start" menu.

Figure

Figure 5.1b Open a command prompt under Windows (XP): using the "Run …" command menu.

Example Execution

In the command prompt, the working directory can be changed to "C:\blast-2.2.23+" by typing "cd \" followed by "cd blast-2.2.23+". If the first prompt is a drive other than "C:\", "C:" instead of "cd \" should be used first. Figure 5.2 contains example commands and their console output for a work session that tests a blast-2.3.23+ installation.

Figure

Figure 5.2 The output of a work session testing the blast+ installation.
The command inputs are in red boxes. The test files produced by the blastdbcmd and blastn command inputs are marked by red arrows.

blastcmd –db refseq_rna –entry nm_000249 –outform "%f" –out test_query.txt

blastn –query text_query.txt –db refseq_rna –out output.txt

The original standalone BLAST package based on NCBI C-toolkit (legacy blast) is deprecated. The installation of legacy blast package for Windows differs from that for blast+ described above. The key differences are summarized below.

a.

The legacy blast packages are located under a different ftp directory:

" ftp://ftp.ncbi.nlm.nih.gov/blast/executables/release/LATEST/

b.

The packages are named with this convention: blast-#.#.#-CHIP-win#.exe, where #.#.# is the version, CHIP is the chipset, and win# is the operating system (32 or 64 bits)

c.

The packages do not contain installer function. It is recommended that the downloaded package be placed in a folder named blast-#.#.# (#.#.# to indicate version) first before extraction

d.

Double clicking the package will execute the self-exacting function to install the package by re-creating bin, doc and data subdirectories

bl2seq.exe     blastall.exe     blastclust.exe     blastpgp.exe
copymat.exe    fastacmd.exe     formatrpsdb.exe    implala.exe makemat.exe
megablast.exe  rpsblast.exe     seedtop.ext

e.

The programs under the bin subdirectory have names and functions different from that provided by blast+: Input: List of UIDs (&id); Source Entrez database (&dbfrom); Destination Entrez database (&db)

f.

Configuring the legacy blast installation is similar to blast+. However, a additional DATA environment variable with the path to the "data" subdirectory as its value should be specified

g.

If the refseq_rna database mentioned in section 3 is installed, the following command lines can be used to test the installationOutput: XML containing linked UIDs from source and destination databases

Questions, feedbacks, and technical assistance requests should be sent to blast-help at:

" blast-help@ncbi.nlm.nih.gov

Questions on other NCBI resources should be addressed to NCBI Service Desk at:

" info@ncbi.nlm.nih.gov