Characterization and similarity analysis of DNA sequences grounded on a 2-D graphical representation

Based on the classification of the four nucleic acid bases, we reduce a DNA primary sequence to three (0, 1) sequences: non-A, non-G and non-C. We associate each sequence with a matrix by giving a 2-D graphical representation, and thus obtain a 3-component vector with entries being the sum of the maximal and minimal eigenvalues of the matrices. The introduced vector is applied to characterize and compare the coding sequences of exon 1 of beta-globin gene of 11 different species.