There are indeed quite a few multiple alignment algorithms. Wallace et al counted around 50. They started drawing trees of alignment algorithms.
Has anyone collected a list of all papers that developed a new alignment algorithm (each one, of course, better than a couple of the others)? I personally would bet that - given that the number of algorithms raise with algorithmical simplicity and the interpretability of the results - motif discovery on DNA is one of the disciplines that generated the most different papers about a new algorithm (counted around 80). Fortunately, the decision is much simpler as the savvy bio-computerfreak knows: With so much choice, the first program that gently compiles after "make" has a good chance of getting used in the end. Is this the reason why everyone is using BLAST today? Or was that a completely different time?
Comments
Another method paper
Hey hey, look at this, I've found a really cool methods paper: Piva et al 2006 They are describing a tool that (tataaa!) generates random sequences!!! Really! Generated with the Delphi function that generates random numbers. Impressive, isn't it!
Wu & Manber
We used Wu & Manber's algorithm [1], famously used in agrep, to recover a set of known subsequences from a longer sequence. For a set of 4 (one each taken from 4 pools of ~150 candidates, each of ~20bp length), we can recover the 4 members in each of 10,000 sequences in about 2 minutes on a standard desktop PC. The algorithm also matches with a specified error count - which can be handy if the sequencer is playing up!
[1] S. Wu and U. Manber, ``Fast Text Searching Allowing Errors," Communications of the ACM 35 (October 1992), pp. 83-91.
Recursive bioinformatics
How can I trust that tree without a tree of methods for building trees?
EMBOSS lives again
Guys the best known open source bioinformatics suite EMBOSS [ http://emboss.sourceforge.net/ ] get a new life. "The funding from the BBSRC means that EMBOSS co-founders Peter Rice and Alan Bleasby will be able to continue the EMBOSS project at the EMBL -EBI for the next three years." [ http://www.genomicsproteomics.com/index.aspx?ID=71972 ].
EMBOSS has almost all the well known bioinformatics algorithms in separate programs as well as API's.
I just love this program and I am very happy to hear this news.
Ani
______________________"The Answer Lies in Genome"______________________
http://fuzzylife.org/animesh/
good emboss news
Pleased to hear it. I'm a big fan of EMBOSS too, especially in combination with BioPerl's Bio::Factory::EMBOSS. It seems to have been chugging along OK as a sourceforge project but I'm sure the funding will help a lot.
New additions
Waiting to see what new stuff they add to the suite. I am waiting to see things on systems biology and networks.
Tree of methods?
A tree of alignment algorithms. Now I've seen everything.
One of my pet hates in bioinformatics is excessive method development. I appreciate new tools, algorithm development and so on, but there are far too many papers reporting tiny incremental improvements in pre-existing methods. People coming in from compsci need to remember what biologists want. They want a tool that they can understand, have some idea of its pros, cons, accuracy and apply to biological data. "I want action - not theory!" as a sample proclaims on a "Man...or Astroman?" record.
phylogeny programs
I should have added joe felsentein's list of phylogeny programs to that post. quote: "Here are 267 of the phylogeny packages, and 32 free servers, that I know about." I admit that my example motif discovery is nothing against old disciplines like that...