We transform primary protein sequences into condensed 20-tuple mathematical descriptors, which are based on the singular values decomposition of the matrix mapped from the original amino sequence. The extracted 20-D condensed feature vectors (CFV) facilitate our quantitative analysis of protein sequences further. Using the condensed representation of the primary protein sequences, we analyze the similarity of nine species based on their ND5 sequences. We also compare the results in this study with those of other related work. (C) 2012 Elsevier B.V. All rights reserved.