Project Description

SARS-CoV-2 is a member of a large family of viruses called coronaviruses that caused a tremendous danger to the worldwide pandemic, leading to millions of deaths. As of May 31, 2021, the WHO has proposed labeling the SAR-CoV-2 variants and the scientific terminology using Greek Symbols. Based on the increase in transmissibility and infection severity, the variants have been classified into Variants of Concern (VOC) and Variants Of Interest (VOI). WHO has categorized four variants as VOC, and eight variants come under VOI. Further, a predictor is highly required to distinguish strain types with the use of their genomic information. To mitigate this problem, we first proposed the framework of deep learning to diagnose the COVID19 based on simple clinical signs and symptoms using a gradient-boosting machine model built with decision-tree base-learners. Then 1D conventional neural network (CNN) will be used to predict the SARS-CoV-2 strain type from the genomic information of viruses underlying prediction with an alignment-free technique. Finally, to verify the efficacy of predictor, it has to be compared with other state-of-the-art prediction techniques based on Linear Discriminant Analysis, Random Forests, and Gradient Boosting Method.