Identification of Novel Protein Sequencing SARS CoV-2 Coronavirus Using Machine Learning

ABSTRACT

The World Health Organization (WHO) declared Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection as a global pandemic in March 2020 causing COVID-19 (Coronavirus Disease-19). Till date, more than 173 million people have been infected worldwide, whereas more than 3.7 million deaths have already been reported caused by COVID-19. Protein-to-protein (PPI) interaction plays an important role in the cellular process of SARS-CoV-2 virus infection in the human body. Although the recent emergence of SARS-CoV-2 has prompted a push for deeper understanding of SARS-CoV-2 and development of effective treatment. However, understanding SARS-CoV-2 is even more critical. It was previously discovered that the proteome of the virus was known, and thus it was possible to derive some of the protein structures by experimentation and others by model-based prediction approaches. The results are later verified by experiments. Considerable research attention has been directed toward DEE deploy features extraction algorithm, amino acid composition AAC and pseudo amino acid composition PseAAC algorithms. We have proposed AdaBoost classification models and compared them with other two machine learning classifiers, such as K-Nearest Neighbor and Random Forest. This paper is not intended to be a comprehensive evaluation of AdaBoost, K-Nearest Neighbor, and Random Forest, rather we have used these models to create an ensemble classifier with excellent performance metrics such as accuracy, precision, specificity, recall, and F1 score. Based on the ensemble model, 1326 total human target proteins are predicted to be potential SARS-CoV-2 viral proteins.