[Biodevelopers] Needleman-Wunsch

[Boris Steipe]

Absolutely! The probability of introducing a gap at a given position  
is much smaller (~10 times) than the probability that the gap will  
include more than just a single amino acid. Gaps do not grow by  
linear addition of amino acids.

The problem is that we do not have good physico-chemical models how  
indels are generated, just empirical observations. These seem to  
indicate that an exponential extension term would be better than a  
linear one, but it is computationally more costly.

There is literature on that, which you should be able to find ... for  
example Gaston Gonnet's 1992 study (newer studies would probably cite  
that one so you can find them through the citation index):

Gonnet GH, Cohen MA, Benner SA.
Exhaustive matching of the entire protein sequence database.
Science. 1992 Jun 5;256(5062):1443-5.

HTH



Boris



On 14 Jul 2006, at 09:58, Theodore H. Smith wrote:

>
> Do bio-informaticians normally expect a Needleman-Wunsch  
> implementation, to allow for differing gap open and gap extension  
> penalties?
>
> Or is it normal to just have a linear gap penalty.
> _______________________________________________
> Biodevelopers mailing list
> Biodevelopers at bioinformatics.org
> https://bioinformatics.org/mailman/listinfo/biodevelopers