Deep learning in managing SARS-CoV-2 pandemic, variant prediction and driver mutation identification: Case study in Setif, Algeria

The COVID-19 caused damage across the globe since it first surfaced in 2019, and even after the worldwide health emergency, researchers are still working hard to comprehend the mechanisms underlying Severe acute respiratory syndrome SARS-CoV-2 mutation and transmission. This article presents a molecular screening study conducted on SARS-CoV-2 ribonucleic acid (RNA) extracts in the wilaya of Setif, Algeria, highlighting the critical role of bioinformatics and machine learning in addressing the COVID-19 pandemic. By leveraging advanced computational techniques, the study identifies mutations and variants responsible for epidemic peaks, offering valuable insights for future predictions and therapeutic interventions. The research employs various machine learning models, including deep multilayer perceptrone (MLP), k-nearest neighbours (KNN), and support vector machine (SVM), to detect Covid-19 variants and mutations with varying degrees of accuracy and precision. Deep MLP demonstrates promising performance in identifying both variants and mutations, as well as its significant role in managing health care pandemic and tragedies.