Article Preview
Top1. Introduction
The increasing probability of death overtime related to the characteristic change in the composition are obtained by aging. Aging (Akiguchi et al., 2017; Onodera et al., 2000), aconnected-time disintegrate of physiological purpose, are obtained by the organism’s cell, tissue and organs, leading (Chen et al., 2019; Sheela & Rangarajan, 2018) to replace such as coloration of skin, organ senescence and skin wrinkling. Aging (Fattah, Lafta, & Alrashid, 2020) itself cannot be considered as illness but poses risk for the practical decay of many organs. Particularly, aging is at the center of age-connected illness such as insanity. SAMP8 strains are the model for super brain-aging of learning and defects of memory. The hippocampus is a crucial part of the visceral brain, precisely governing feeling and noesis (Chen et al., 2019).
To specify insights which means improving the psychological features for reversal of aging in the S8 strain, The structure of vegetative cells in the hippocampal region were additionally examine (Chen et al., 2019). The initial indicators for aging must be the gene expression between the senescence-resistant and senescence-prone strains or it must be indicating accelerated senescence within SAMP8 and SAMP10 strains. There are many dropped genes in the metaphysics of the fascinating cistron (GO) class as concluded by the enrichment analysis of cistron victimization of the GO (Sheela & Rangarajan, 2018). For each of the experimental procedures, the performance of real-world and analytical methodologies was compared to determine organic phenomenon changes; genes were selected from the analysis of the hippocampus (from humans not employed in the analysis of microarray). Of the 10 genes examined, the modifications of zfgh7 expression for genes were fixed with a change which is bigger than three-fold (Sheela & Rangarajan, 2018). To build a speed model with the highest accuracy for predicting the behavior gene in both retina and hippocampus the authors faced challenges in computational biology (Sheela & Rangarajan, 2018). Based on the importance and lack of a definite treatment for the disease, a new technology by using microarray has been applied to define the genes that causes the disease. Microarray technologies are significant tools at medical that the biologist use to monitor gene expression levels in a given organism (Chen et al., 2019). However, the data of gene expression produced from the technology of microarray can cause problems with the methods of classification because the genes number in the microarray data is huge whereas the samples number is very small. In data mining, this simple truth is defined as the dimensionality curse. One of the efficient techniques to address the dimensionality curse is the gene selection, which selects relevant and informative genes.
In fact, gene selection is the process by which a subset of informative genes is selected from the original dataset. This subset of genes helps researchers to gain significant insight a biological structure for the disease (Alanni et al., 2019).
In this study, gene selection based on ANOVA method were used to select the informative genes and enhance the performance of classification. Thus, gene selection method improves the accuracy of classification (Zeebaree et al., 2018)(Alrashid, 2020). Then, CNN model is employed to predict behaviour of the Senescence-Accelerated Mouse (SAM) Strains (SAMPs and SAMR) using gene expression data and is improved predictive performance. The novelty of this study lies within the combination of the gene selection methods and the classifier model to achieve the best accuracy and training times compared to other approaches suggested by other studies (Fattah et al., 2020). In (Zeebaree et al., 2018), the authors proposed a deep learning method via the Convolutional Neural Network (CNN), to classify the microarray dataset. The experiments are implemented on ten cancer datasets. The experimental results showed that the proposed system has the ability to reduce the genes and enhance the performance of cancer classification (Zeebaree et al., 2018).