Microarray based Geonomic Biomarker Optimization for Cancer Prognosis

• Published in: 2023 Second International Conference on Electronics and Renewable Systems (ICEARS) pp. 478 - 483
• Main Authors: K, Jose Triny, S, Kesavan, H, Navadeepan, P, Ranjith Kumar
• Format: Conference Proceeding
• Language: English
• Published: IEEE 02.03.2023
• Subjects: Classification algorithms; Convolutional Neural Network Algorithm; Disease Prediction; DNA; Gene expression; Median Estimation; Particle Swarm Optimization Algorithm; Prediction algorithms; RNA; Sequential analysis; Severity Analysis; Single-cell Ribonucleic acid sequencing; Skin
• DOI: 10.1109/ICEARS56392.2023.10084989

Scientists can now assess the expression of hundreds of genes concurrently in a single experiment because to the advancements in DNA microarray technology. Gene expression patterns, which depict the molecular state of a cell, offer immense potential as a diagnostic tool. The solutions to resolving the fundamental harms associated with diagnosis and discovery is known to be found in the categorization of diseases transcriptional activation utilizing data. Complete simultaneous monitoring of several gene expressions is now achievable because to the recent development of the DNA microarray technology. The possibility of disease diagnosis utilizing gene expression data has begun to be explored by specialists thanks to the abundance of gene expression data. The previous years, a sizable number of procedures there have planned with positive outcomes. Embedded techniques have the prevalence advantage of interacting with the classification algorithms; while on the same time have decrease computational time compared to wrapper strategies. A deeper examination of the issue, the suggested remedies, and the related problems as a whole is required to develop understanding of the disease categorization challenge. In this study, we present a thorough searching method, clustering method, as well as classification techniques including such Trend similarity such as algorithm for particle swarm optimization and evaluate them predicated on their review time, classification accuracy, and capacity to reveal biologically significant gene information. Depending on our classification method, we diagnose conditions including cancer (of the lung, heart, breast, and skin) and other conditions. We also determine the severity of conditions and recommend medications for those conditions using deep learning algorithms. Our test results demonstrate better accuracy in classifier performance when used with graphs.