AUTHOR=Benarjee Shaik , Kumar Vaegae Naveen 

TITLE=A computational approach for prediction of exons using static encoding methods, digital filter and windowing technique

JOURNAL=Frontiers in Signal Processing

VOLUME=Volume 5 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/signal-processing/articles/10.3389/frsip.2025.1679555

DOI=10.3389/frsip.2025.1679555

ISSN=2673-8198

ABSTRACT=IntroductionIdentifying protein-coding regions in eukaryotic Deoxyribonucleic acid (DNA) remains difficult due to the sparse and uneven distribution of exons.MethodsThis work focusses into four static encoding schemes—integer, Voss, paired numeric, and Electron-Ion Interaction Potential (EIIP) to improve exon prediction using genomic signal processing. Two benchmark sequences, Caenorhabditis elegans Cosmid F56F11.4 and Mouse apolipoprotein A-IV (M13966.1), were analyzed in MATLAB. A Cauer (elliptic) band-pass filter was used to isolate the period-3 component, and a Blackman-Harris window was utilised to reduce spectral leakage. The elliptic filter in conjunction with EIIP-based encoding achieved the most distinct separation between coding and non-coding areas among the assessed techniques, identifying every exon segment with a minimal amount of noise.Results and discussionThe technique obtained 84% sensitivity, 96% specificity, and 94% accuracy on the C. elegans Cosmid sequence and 86.5% sensitivity, 93% specificity, and 91% accuracy on the M13966.1 gene sequence.ConclusionThese results show that the EIIP, Cauer filter and Blackman-Harris windowing framework offers a reliable and effective method for identifying exons.