Use the Smith-Waterman-Algorithm for (local) fuzzy string match

Hello,

I already discovered the Levenshtein file on mathworks.com ( http://www.mathworks.de/matlabcentral/fileexchange/17585-calculation-of-distance-between-strings) but I need to adjust the penalty at the beginning and end of the string so I get a locally string match instead of a global one.

Example:

string1 = 'blabla football blabla'

string2 = 'football'

• Levenshtein gives me a score of 14
• Smith-Waterman a score of 0

So does somebody now how to do this?

I have the bioinformatics toolbox as well and there the algorithm is already implemented, but works only for a given set of characters as it is designed for nucleotid (protein) sequences like 'GTCAA', thus gives me an error in this case.

Help is greatly appreciated!

Regards

Answer 1

[Léon]

No ideas? :-(