hi,

I'm working in sequence alignment with non biological sequences.

 

I read much about gap penlty, each author deal with gaps differently.

how i can decide which policy is best?

 

 

 

 

 

thanks, huda

 

 

What algorithm are you using? It's a judgement call based on how many mismatches vs gaps you want in the alignment. What I would probably do is try running the algorithm over a range of gap penalties and calculating some metric (ave gaps and/or ave mismatch). Graph the metrics versus gap penalty and perhaps they'll be some inflection point where the tradeoff is acceptable. That's sort of the quick and dirty way for parameter 'tuning'