CpG islands identification in DNA sequences using modified P-spectrum based algorithm

Abstract

The identification of CpG Islands play a major role in the analysis of DNA sequences because of association of CpG Islands with many epigenetic events. Some of these events are promoter activity and consequently gene prediction, chromosome inactivation, and for early detection of cancer etc. And hence the exact identification of CpG Islands in the stretch of DNA sequences has always remained a challenging task. Numerous computational algorithms have been developed for the identification of CpG Islands in DNA sequences. Recently various transform based methods have been reported for the CpG Islands detection in literature. In these transform based methods, there is a requirement of transforming the signal from time domain to frequency domain and correspondingly there is probability of transform biasing. Hence to overcome this issue, a modified P-spectrum based algorithm has been proposed in this paper. Also the performance of the proposed method has been compared with recently reported methods of CpG Islands detection using standard evaluation metrics. The performance of proposed method has been proved to be much better than the other methods and hence the proposed approach is an efficient method in detecting CpG Islands.