Abstract: According to a mapping of codons and amino acids and stop signal， the sequence of pseudo amino acid for DNA sequence was proposed. Then， by means of the multiset， a 21\|dimensional numerical vector of a DNA sequence was constructed. On the basis of the vector， the similarity distance between any two DNA sequences can be calculated. The phylogenetic analysis on three datasets （S segment of hantaviruses， complete genome sequences of Tomato yellow leaf curl virus and complete genome sequences of human rhinovirus） demonstrated the effectiveness of the proposed method.

Key words: DNA, pseudo amino acid, multiset, phylogenetic analysis