ClustalW and Omega of EBI and Blast of NCBI, both globally align sequences. Smith waterman of EBI aligns sequences locally, but works with just two sequences. How can I align more than 2 sequences locally?
You have to do it pairwise (there is no other option).
You can do that with
BLAST or even
SSEARCH (Smith-Watermann). If you have a
fasta file of your sequences then do this:
- make index of your sequences using
makeblastdb -dbtype nucl -in <yourfastafile> -out <yourdbindex>
<yourfastafile>as input and
<yourdbindex>as database with desired parameters. This will do a pairwise alignment between all sequences
Similar strategy can be used for
Even Water (in EBI itself) is Local alignment tool for multiple sequence alignment. Also in BLAST you get bla2seq which is also used for pairwise alignment. In any of these tools you just have to enter FASTA sequences (any number you want) or any sequence with > sign and run it against the subject (database or your required sequence). If you want to see the results for all the alignment in one screen, a tool like BioEdit would be ideal. I hope this is helpful, I'm not an expert in bioinformatics so feel free to correct me :)
I use Jalview, that is a small program written in Java. It is simple to use and can do many many things. It's web-service based and it has 7 different services built in just for sequence alignment. You can load your sequences from many file formats, or get them by reference id. It's really a small swiss-army-knife. I really recommend to take a look at it. It's available on http://www.jalview.org/.
I might have missed the local alignment part earlier, but I did some research and found that MULAN (MUltiple sequence Local AligNment and conservation visualization tool ) may be what you're looking for.
protected by Community♦ Mar 18 '15 at 12:24
Thank you for your interest in this question.
Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).
Would you like to answer one of these unanswered questions instead?