Abstract
In this study, we evaluated the use of a predictive computational approach for SARS-CoV-2 genetic variations analysis in improving the current variant labeling system. First, we reviewed the basis of the system developed by the World Health Organization (WHO) for the labeling of SARS-CoV-2 genetic variants and the derivative adapted by the United States Centers for Disease Control and Prevention (CDC). Both labeling systems are based on the virus' major attributes. However, we found that the labeling criteria of the SARS-CoV-2 variants derived from these attributes are not accurately defined and are used differently by the two agencies. Consequently, discrepancies exist between the labels given by WHO and the CDC to the same variants. Our observations suggest that giving the variant of concern (VOC) label to a new variant is premature and might not be appropriate. Therefore, we used a comparative computational approach to predict the effects of the mutations on the virus structure and functions of five VOCs. By linking these data to the criteria used by WHO/CDC for variant labeling, we ascertained that a predictive computational comparative approach of the genetic variations is a good way for rapid and more accurate labeling of SARS-CoV-2 variants. We propose to label all emergent variants, variant under monitoring or variant being monitored (VUM/VBM), and to carry out computational predictive studies with thorough comparison to existing variants, upon which more appropriate and informative labels can be attributed. Furthermore, harmonization of the variant labeling system would be globally beneficial to communicate about and fight the COVID-19 pandemic.