This chapter contains protocols for the most common BLAST searches. Because every BLAST experiment is unique, you should treat the protocols as a starting point and use your own knowledge about BLAST to modify the procedures. The discussions include what to do, as well as why. Although this approach makes the descriptions more verbose, explaining the logic behind these choices will help you make intelligent choices when creating your own protocols.
Most BLAST experiments fall into one of two categories: mapping and exploring. Mapping is the process of finding the position of one sequence within another?for example, finding a gene within a genome. When mapping, you can expect the alignments to be nearly identical, and the coordinates are generally the focus of the results. When exploring, the goal is usually to find functionally related sequences. When exploring, your alignment statistics (score, expectation, percent identity, etc.) are often of greatest importance, at least initially. Making functional and phylogenetic inferences, especially between distantly related sequences, often requires inspecting the alignments from a biological rather than a statistical perspective. There is, of course, a continuum between mapping and exploring, but keeping this dichotomy in mind can help you zero in on the fundamental aspects of a search strategy.
The notation used here and in the reference chapters in Section 5 is the command-line interface. If you're unfamiliar with shells and terminals, the command line is where you type in program names and options. It may seem a little odd at first, but it is analogous to filling in a web form and then clicking the submit button. While most people use BLAST via some web interface, not all pages look the same or support the same parameters. Behind the scenes, though, they all interact with a command-line version of BLAST. BLAST pages frequently let you set advanced options; usually it's a text box or boxes in which you can enter the command-line options.