Extended Data Fig. 2: Comparing the probability of emergence of the vaccine resistant mutant for different model extensions.

For each of the model extensions presented in Supplementary Fig 4 (social distancing adjusted every week), 5 (seasonality) and 13 (vaccine hesitancy), we ran 1000 independent runs and reported the calculated probability of emergence of the vaccine resistant mutant. For some parameter sets (for example L = 100, c = 10,000), we did not observe any takeover events out of 1000 runs of the simulation for the basic model, social distancing every week extension and seasonality with 10% decrease extension. We did not observe any significant differences when incorporating any of the following extensions: (i) social distancing every week instead of every day; (ii) low level vaccine hesitancy; (iii) 10% increase in infectivity in winter; (iv) 10% decrease in infectivity in winter. Parameters: N = 10⁶; a = 0.25; d = 0.01; μ = 10⁻⁶; s₀ = 0.1, β₁ = β₂ = 7.5 ∙ 10⁻⁷. Error bars are calculated as in Extended Data Figure 2, and the number of runs is n=1000.