In computational biology, DNA sequence analysis holds significant importance. Initially, sequence alignment methods were employed for this purpose. Subsequently, sequence alignment free methods were employed, utilizing distance as a metric. These methods measure similarity based on short distances for similar sequences and long distances for dissimilar ones. This paper introduces a novel approach to DNA sequence analysis, this approach needs extracting features from the DNA sequences. Subsequently, a comparison is made between the features of one sequence against another to identify common and distinctive features. The similarity is then defined as the ratio of common features to the union of common and distinctive features. Consequently, a higher degree of similarity is inferred when more common features are shared between two sequences, while a greater number of distinctive features suggests greater dissimilarity. Furthermore, the proposed approach is compared against various coefficients that evaluate similarity based on the presence or absence of specific data. Ultimately, the comparative analysis concludes by noting similarities between the proposed approach and several of these coefficients.