Detection and Localization of Hidden Patterns in DNA Sequences Using Signal Processing

Abstract

newline The completion of human genome sequencing project in April 2003 and subsequently next generation sequencing technology provided the direction for annotation of genome to come into existence. The huge amount of genomic raw DNA data is annotated in genome annotation by extracting useful information and the annotated data is added to the data base of genome. Genomic data available in the form of DNA sequences consists of various hidden patterns which are associated with the functioning of the organism. Protein-coding regions, introns, CpG Islands, tandem repeats, genic regions, inter genic regions, promoter regions, transcription start sites, and untranslated regions are few examples of sections of DNA which are important. Many computational approaches have been developed for the identification of these regions. However, development of accurate and efficient approaches for the detection and localization of the hidden patterns in DNA sequences has always been a challenging task. In this research work, efficient approaches have been proposed for the detection of hidden patterns such as protein-coding regions, CpG Islands, and tandem repeats in the DNA sequences. The signal processing based tools have been employed in all the proposed approaches of this research work. The platform used for the simulation of proposed algorithms in this work is MATLAB (R2013). The performance assessment has been carried out using standard evaluation metrics and the comparison has been done with recent state-of-art methods on the benchmark datasets. The proposed approaches have achieved significant improvement in detection over recent state-of-art methods.