A time series forecasting of the proportion of SARS-CoV-2 N501Y lineage in North America

Abstract

Introduction: The outbreak of pneumonia known as SARS-COV-2 and newly-emerging South African (B.1.351), the United Kingdom (B.1.1.7) and Brazil (P.1) variants have led to a more infectious virus and potentially more substantial loss of neutralizing activity by natural infection or vaccine-elicited antibodies. Methods: We identified prevalent mutations using the spike receptor-binding domain (S-RBD) of SARS-CoV-2 deposited in the Nextstrain global database and comparing them to the Wuhan-Hu-1/2019 genomic sequence as a reference. Then we calculated the percentages of mutant genomes from the total regional subsample isolates from December 2019 to the end of January 2021. We developed two separate time series forecasting models for the SARS-CoV-2 B.1.1.7 variant. The computational model used the structure of the S-RBD to examine its interactions with the neutralizing antibody, named CV30 (isolated from a patient), and human angiotensin-converting enzyme 2 (hACE-2), based on a hybrid algorithm of template-based modeling to predict the affinity of S protein to the neutralizing antibodies and hACE-2 receptor. Results: The proportion of the B.1.1.7 strain in North America is growing fast. From these computations, it seems that the S-RBD and hACE-2 proteins are less favorable for the South African strain (K417N, E484K, and N501Y) as compared to the wild type structure and more favorable for B.1.1.7 and P.1 variants. In the present of crystallized CV30 neutralizing antibodies, docking scores suggest antibodies can be partially neutralize the B.1.1.7 variant, and, less efficiently, the B.1.351 and P.1 variants. Conclusion: The rapid evolution of SARS-CoV-2 has the potential to allow the newly-emerged B.1.351, and P.1 variants to escape from natural or vaccine-induced neutralizing immunity and viral spreading.