Cloud Based Exon Prediction Methodology using Logarithmic Adaptive Algorithms for Genomic Signal Analysis

Abstract

Cloud Computing provides healthcare companies with the major convenience in the research and economic aspects. Cloud services ensure that large quantities of such sensitive data will be stored and managed securely. The gene sequence labs send inferred data to various sequence libraries under traditional flow of gene information through the Internet. Cloud service use will reduce DNA sequencing storage costs to a minimum. In this work, we developed a novel genomic bioinformatics-based system, using Amazon Cloud Services, that stores and processes genomic sequence information. An important task in bio-informatics that is used for the recognition and implementation of disease drug, is the exact recognition of exon regions in deoxyribonucleic acid (DNA) sequence. All exon identification techniques are based on three basic periodicity (TBP) properties of exons. With the comparison of various existing methods, adaptive signal processing techniques have been promising. This paper uses the maximum normalized logarithmic mean least square (MNLMLS) algorithm also its signed variants to develop multiple adaptive exon predictor (AEPs) with less computational complexity. Eventually, a performance evaluation is performed for different AEPs using various standard gene data sequences derived from National Biotechnology Information Centre (NBI) genomic sequence database, such as Sensitivity (Sn), Specificity (Sp) and Precision (Pr) measurements.