Reasons:
1. Comparing M1(p*q) and M2 (r*s) will generate p*r score matrices, each SM one Needleman-Wunsch algo, and p*r NW algo applied on these SM's. Totally 2*p*r NW algo, which is too computationally costly.
2. Distance matrices of proteins are symmetric; dealing with asymmetric matrices have to care about the transposed matrix.
No comments:
Post a Comment