Adaptive Exon Prediction for Disease Identification using Higher Order Algorithms

Imperative part in bio-informatics is to identify protein-coding sections in deoxyribonucleic acid (DNA) sequence using smart systems based on adaptive methods. Exon area analysis remains important for disease detection plus to develop drugs. In exon areas, three base periodicities (TBP) are noted. Adaptive techniques present unique ability to modify genome sequence in dependent coefficients of weight. A critical task in bio-informatics, which helps in disease diagnosis and accurately locating exon segments in DNA. All exon segments in DNA exhibits the property of three base periodicity (TBP) that helps in locating them precisely. From these, we introduce a novel adaptive exon predictor (AEP) with a Normalized Least Mean Fourth (NLMF) also its signed variants NSRLMF, NSLMF, NSSLMF to lower computational complexity when compared to LMS. It is shown that AEP based on SRNLMF is better depending on performance metrics like precision 0.6789, sensitivity 0.7129, also specificity 0.7343 at threshold value of 0.8. From National Center for Biotechnology Information (NCBI) gene repository, genomic sequence of homo sapiens are taken by using AEPs that are able to predict exon location.