when the local alignment tends to work as global alignment?

Question

[jafarzadeh91]

Hi

I read somewhere that the local alignment between two sequences of the roughly same length tends to work as global alignment when the substitution and gap score is big. But It didn't explain why.I would be appreciate if you can provide me an intuition or a proof of the reason behind that.

Answer 1

[Brian Bushnell]

The lower the penalty is for gaps and substitutions, the closer a local alignment will approach a global alignment. High penalties will increase the difference by encouraging clipping. The relative length of the two sequences does not matter.

Answer 2

[Biomonika (Noolean)]

If gap score is high, then it is very costly to open the new gap and the algorithm will prefer to rather tolerate stretch of mismatches when progressing alignment and introducing gap only when absolutely necessary. Hence, instead of reporting short region with perfect local alignment, it will be forced to align complete sequences, just like in the global alignment. I am not sure about substitution score you mention - if the sequences differ, then really the only options to span those regions are either via mismatches or gaps. I suggest you find some implementation of NW and SW algorithms online, play with the options and visually compare alignments you get.