Species homology analysis method based on amino acid location and physicochemical properties

The similarity analysis of protein sequences is a method for analyzing the homology between species. In this paper, we first encode a protein sequence into a 9 dimensional feature-vector consisting of a 20 dimensional content ratio vector, a 20 dimensional positionatio vector of the amino acids, and a 9 dimensional mean-physicochemical properties vector. Using the Euclidean distance to characterize the similarity distance of the protein sequence. We tested our method on two datasets: (1) 9 species ofND5 sequences, (2) 28 species of influenza A virus sequences. As a result, to verify the validity and practicability, we compare the correlation between the corresponding method and ClustalW.