Alignment-free comparison is a recently developed method for sequence alignment, which has high computational efficiency and suitable to the low identical sequences. Alignment-free comparison was successfully applied in the DNA analysis. However, the accuracy of analysis is not high when it was applied in protein analysis because the complexity of protein is larger than DNA by consisting of 20 types of residues. Thus, residues are clustered into a few groups based on their similarity of physicochemical features. Using such simplified alphabets, the complexity of protein sequences is reduced and at the same time the key information encoded in the sequences remains. Therefore, the accuracy of alignment-free comparison is improved.