On spectral numerical characterizations of DNA sequences based on Hamori curve representation
We present new numerical characterization of DNA sequences that is based on the modified graphical representation proposed by Hamori. While Hamori embeds the sequence into Euclidean space, we use analogous embedding into the stron gproduct of graphs, $K_4 \boxtimes P_n$, with weighted edges. Based on this representation, a novel numerical characterization was proposed in [Pesek and Zerovnik, MATCH Commun. Math. Comput. Chem. (Krag.), vol 60, 2008, 301-312] which is based on the products of ten eigenvalues from the start and the end of the descending ordered list of the eigenvalues of the $L/L$ matrices associated with DNA. In this paper we compare two further numerical characterizations of the same type emphasizing the robustness of the approach.