A novel DNA mapping scheme for improved exon prediction using digital filters

Abstract

It has become a challenging issue for the researchers to predict protein coding regions in context of Signal Processing and bioinformatics. This paper presents a novel mapping technique and a hybrid approach combining Gauss windowing and FFT techniques for improved gene prediction. Mapping scheme plays an important role in accuracy of exon prediction. Some mapping scheme works well in a particular gene set but may not perform better in some other gene predictions. The three base periodicity properties is used to identify the exons location. In addition to this a threshold formulation has been used to investigate the performance of the algorithm in terms of various parameters like accuracy, sensitivity, specificity, geometric-mean and execution time. Analysis is performed in both nucleotides as well as exon level detection. The performance of the proposed mapping is compared with numerous existing mappings and also tested on NCBI genes which prove the efficiency of the algorithm.