Database searches

Editorial: These programs search one sequence against a database of sequences. Blast is the most popular by far of these, though FastA is a slower alternative that is often more sensitive and produces better alignments. HMMer is a little different in that is uses Hidden Markov Model profiles of a family of sequences in the search; this is most often used to find protein domains, such as the HMMs of the Pfam or SMART collections.

The swiss-army tool of sequence bioinformatics, used to compare one protein or DNA sequence against a database of sequences, for a myriad of reasons. The most popular version comes from NCBI (look for the file 'blast[version]-ppc32.macos.tar.gz'). You can enhance this package by replacing the main blastall program with a version optimized for Mac, by Apple and Genentech (FTP).

The standard version of Blast uses a text (command line) interface, but NCBI also makes a version that runs through a web server, even on your own machine. I haven't tried their Mac version yet, but if you wish to give it a try, it's in the same directory as above, the file starts with the text 'wwwblast'.

An alternative to NCBI Blast is WU-Blast, which claims to be faster and more sensitive. MacOS X versions are available under licence (free for academics) from the above page. The uncompled 1.4 version is also available freely.

Generally slower but more sensitive than Blast, the FastA package can align individual pairs of sequences or run a sequence against a database, including the rigorous Smith-Waterman alignment method. Command line interface only. The BioTeam have provided a copy of FastA 3.3 for Mac OS X. See also FastA documentation.

For analysis of protein families and domains, the HMMer package will compare profile HMMs (a composite of many related sequences aligned together) with individual sequences. HMMer is created by Sean Eddy, and the Mac-optimised port is by Erik Lindahl.