The Three Base Periodicity of Protein Coding Sequences and its Application in Exon Prediction

Abstract

periodicities within the genomic sequences play an important role in understanding the structural and functional behavior of DNA signals. The three-base periodicity (TBP) inherent to protein-coding regions is used as a discriminative feature for locating exons in eukaryotic DNA sequences. DSP tools such as Short time discrete Fourier transform (STDFT) and anti-notch filter (ANF) based techniques exploit three base periodicity property in spectral domain to distinguish exons and introns within a gene. In the last two decades, several other approaches have been evolved that are also based on this particular feature. However, the performance of these techniques is still not satisfactory. In order to enhance the prediction performance of these techniques, it is important to understand the underlying concepts of tree base periodicity and its impact on exon prediction. In this work, rather than proposing any new feature extraction technique, we have discussed the concept of the 3-base periodicity and analyzed its impact on DSP based exon prediction techniques. It is obtained that the period-3 feature is less prominent in the short length exons and therefore it is particularly difficult to predict. Moreover, the comparative study of STDFT and ANF technique is also presented.