In biological information systems, the analysis of biological sequences is a major problem in today's bioinformatics research. In this paper, a method for protein sequence similarity analysis based on discrete wavelet transform and fractal dimension and single clustering method is proposed. Multivariate decomposition of digital signals containing biological information is performed by discrete wavelets. Using the Higuchi algorithm based on wavelet decomposition, the fractal characteristics of the primary structure of the protein were studied using multiple properties of the protein. The distance matrix between different proteins is obtained by analytical calculation. Phylogenetic tree, and similar analysis of protein sequences. The results show that compared with the traditional methods, wavelet transform and fractal dimension methods and multiple attribute analysis can analyze the similarity of protein sequences more comprehensively, reliably and quickly.