thnx for the reply. A k-mer or k-plet, or k-tuple ...(there are meny deiierent ways people call those). Basically (generaly speaking) it is a k-size constrict of individual elements. these elements are often chosen to be charracters (or letters, or symbols) from an alphabet ... depending upon the literature and the field of interest. Linguists and biologists will usually use the term k-mer (a k character string) (as far as my literature knowlage goes )

As far as your post goes. You perfectly understood the first part of the problem. I want to find all k-mers with score values above 17. And yes I could extract the diagonal and do everything faster. But I need the matrix because then in the second 3 for-loops I want to find all k-mers that , when transhormed into the initial high scoring k-mer the transformation score will not be below 17 (again). that way I am certain that all k-mers with score above 17 are found

the problem is if the k and the alfhabet (6) gets really big, itterating through all possible solutions (like I'm doing) is not feasible. so I'm trying to locate some type of heuristic solution to reduce the number of iterations to minimum .

thnx once more for a suggestiog