Fig. 2
Nonsynonymous (N) and synonymous (S) mutations across frequencies. (A) N/S ratios of mutations at different frequencies. (B) N/S ratio of mutations that occurred 1–20 times in the population. The N/S ratio is highest with mutations that occurred once and rapidly decrease. Significant differences in mutation frequency are only observed between one-time and two-time occurrences, and between two-time and three-time occurrences. (*** p < 10− 3; Chi-square test). (C) Temporal dynamics of SARS-CoV-2 mutations with frequencies ranging from 10⁻³ to 10⁻², classified by their coefficient of variation (CV) over a 16-month period. Red lines represent the top 10% most variable mutations (highest CV), while blue lines indicate the bottom 10% least variable mutations (lowest CV). Each line shows the frequency trajectory of a single mutation. Although mutations in the top 10% display greater fluctuation, neither group exhibits a consistent upward trend, suggesting a lack of clonal expansion. (D) N/S ratios of mutations with frequencies ranging from 10⁻² to 10⁻³, which are significantly higher than those at four-fold degenerate sites, ranked by the coefficient of variation (CV) calculated from their monthly frequencies over a 16-month period. (E) Contribution of genetic diversity (θ) by mutations at different frequency. Watterson estimator of genetic diversity, θ (Watterson 1975; Y-axis), was estimated by randomly selecting 100 SARS-CoV-2 genomes from the dataset. Mutations at different frequency (X-axis) were individually removed to calculate the θ. The process was repeated 100 times and the mean and standard deviation of the θ were estimated. Although mutations within 10− 3– 10− 2 frequency bin only account for ~ 2% of total changes (indicated by an arrow), they represent ~ 37% of genetic diversity