One-dimensional modeling of DNA sequences

Signal processing tools are increasingly being used for DNA analysis, including sequence structure prediction, sequence matching and classification and sequence function identification. Almost all of the proposed techniques perform the DNA analysis in a higher dimensional space, typically of dimension 4 to match the four-letter alphabet {A, C, T, G} representation of DNA sequences. In this study, we show that it is possible to obtain similar results to those reported in the literature by processing DNA data in a 1-D space with lower complexity. We illustrate our approach by focusing on the gene identification problem and comparing the 1-D processing results with prior work.