So "coding" is good for actual ORFs, "maximal" is good for everything else, and "Two templates" just does both?
"discontiguous megablast" on the NCBI website, as well as other implementations, appears to have a "Discontiguous Word Options" parameter set. These include a template length and template type; the latter can be:
- Coding
- Maximal
- Two templates
What do these options do?
1 answer
The way BLAST works is by first matching a word from the query and the database, this match is then expanded using dynamic programming. The word is usually 7-256 in length. When using discontiguous BLAST the word is 11-12 letters in a 25 letters long sequence. The way these letters that must match between the query and database are spread differently along the longer sequence. In coding the matching letters will be in the first two nucleotides of every triplet, in maximal they will be spread in a predefined pattern that should maximize the number of matches, two templates just try to match these two patterns.
In short discontiguous BLAST allows matches that don't have a 7 letters word that perfectly match between the two.
Exactly, although if you're looking for ORFs you should probably use tblastx
Log in to answer this question.