Figure 11a:
Alignment similarities versus substitution matrices (PAM) for different pair wise sequences using N-W algorithms and FIS with sub clustering.