This paper defines the Accumulated Nucleotide Frequency Distribution (ANFD) of DNA sequences. We compute the mean, variance, and covariance of the ANFD for each base, together with the mean of the Accumulated Nucleotide Frequency (ANF). Based on these quantities, a 14-dimensional feature vector is constructed to numerically represent DNA sequences. The Euclidean distance is then employed to quantify the similarity between pairs of DNA sequences. The proposed method is validated experimentally on three DNA sequence datasets. The results demonstrate that the method achieves accurate clustering of DNA sequences, thus verifying its effectiveness.