Abstract
In this paper, we discuss the discrimination of SARS-COV2 viruses associated with three major affected countries the USA, China, and Germany. The discrimination can reveal the mutation as the result of viral transmission and its spread due to mutation associated with its protein structure which makes small changes in the Spike protein. To investigate the mutation in SARS-COV2, we downloaded the protein strains associated with the USA, China, and Germany from the UniProtKB by advance search through SARS-COV2, country name, and protein name: Accessory protein 7b, 6, ORF3a, 10, 8 protein, Envelope small membrane protein, Nucleoprotein, Membrane protein, Spike glycoprotein, 3C-like proteinase, and 2'-O-methyltransferase. After retrieving the protein sequences, we transform the biological form of sequences to their equivalent numerical form by using statistical moments. Further classification algorithms like Random Forest, SVM are used for their training and classification. Finally, performance evaluation is carried out using K-fold cross-validation, independent testing, self-consistency, and jackknife testing. The result received through all testing is more than 97%, which shows the visible discrimination among the protein strains of mentioned countries, which shows the strong mutation in SARS-Cov2 sequences.