The documentation doesn't cover how Ns are handled when evaluating for MAX_POLY_X, but with the current implementation, 'GGGNNN' violates MAX_POLY_X=5. I take this to mean that the MAX_POLY_X evaluation should always be based on the worst possible case (all 3 Ns could be Gs). However, 'GGGNGG' does not currently violate MAX_POLY_X=5 nor does 'NNNNNG' even though substituting Ns for Gs in these examples would result in a run 6 identical base pairs.
The documentation doesn't cover how Ns are handled when evaluating for MAX_POLY_X, but with the current implementation, 'GGGNNN' violates MAX_POLY_X=5. I take this to mean that the MAX_POLY_X evaluation should always be based on the worst possible case (all 3 Ns could be Gs). However, 'GGGNGG' does not currently violate MAX_POLY_X=5 nor does 'NNNNNG' even though substituting Ns for Gs in these examples would result in a run 6 identical base pairs.