ABSTRACT
Much can be learned from 1.2 million sequences of SARS-CoV-2 generated during the last 15 months. Out of the overwhelming number of mutations sampled so far, only few rose to prominence in the viral population. Many of these emerged recently and independently in multiple lineages. Such a textbook example of convergent evolution at the molecular level is not only curiosity but a guide to uncover the basis for adaptive advantage behind these events. Focusing on the extent of the convergent evolution in the spike (S) protein, our report confirms that the most concerning SARS-CoV-2 lineages carry the heaviest burden of convergent S-protein mutations, suggesting their fundamental adaptive advantage. The great majority (21/25) of S-protein sites under convergent evolution tightly cluster in three functional domains; N-terminal domain, receptor-binding domain, and Furin cleavage site. We further show that among the S-protein receptor-binding motif mutations, ACE2 affinity-improving substitutions are favored. While the probed mutation space in the S protein covered all amino-acids reachable by single nucleotide changes, substitutions requiring two nucleotide changes or epistatic mutations of multiple-residues have only recently started to emerge. Unfortunately, despite their convergent emergence and physical association, most of these adaptive mutations and their combinations remain understudied. We aim to promote research of current variants which are currently understudied but may become important in the future.