期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2016
卷号：93
期号：1
出版社：Journal of Theoretical and Applied
摘要：In the field of Bio-informatics, exact predicting the regions that code for proteins in a deoxyribonucleic acid (DNA) sequence is a challenging and vital task. Analyzing the exon regions is a major phenomenon which helps in drug design and disease identification. The sections of DNA that contain protein coding information are known as exons. Hence predicting the exons in a DNA sequence is a crucial task in genomics. Nucleotides serve as the basic structural unit of a DNA. Three base periodicity (TBP) has been practical in the protein coding regions of DNA sequences for nucleotides. By applying Signal processing techniques, TBP can be easily predicted. Adaptive signal processing techniques found to be likely due to their distinct capability, with the ability to change weight coefficients depending on the gene sequence. In this paper, we propose an efficient adaptive exon predictor (AEP) based on these considerations using error normalization. To increase the tracking ability of the adaptive algorithm for exon regions, we develop AEPs using ELMS algorithm and its variants. These proposed AEPs prominently reduces computational complexity and offers superior performance in terms of performance measures like sensitivity, specificity, and precision, so that the AEPs are attractive in nano devices. It was shown that maximum error normalized sign regressor LMS (MESRLMS) based AEP is better in exon prediction applications based on performance measures with Sensitivity 0.7198, Specificity 0.7203 and Precision 0.6906 at a threshold of 0.8. Also, this algorithm performs better with respect to convergence because of error normalization. Computational complexity wise also MESRLMS needs only one multiplication operation because of sign regressor operation and using a maximum value in normalization. Finally the ability of various AEPs in prediction of exons is tested using different DNA sequences obtained from National Center for Biotechnology Information (NCBI) database.
关键词：Adaptive Exon Predictor; Computational Complexity; Deoxyribonucleic Acid; Disease Identification; Exons; Three Base Periodicity
