Figure 3.—

(A) Fitness plotted against substitution number. Each point represents the mean of ≥10 replicate growth assays performed on a single genotype. “Substitution number” is the order in which a mutation that is ultimately fixed appears. The straight line shows the best fit of a linear regression model; the curve is the best fit of a hypergeometric model [y = x₀ + (a × x)/(x + b), where x₀ = 1 and a and b are estimated]. (B) The magnitude of fitness “jumps” plotted against the order in which they occur. The categories on the x-axis indicate which fitness jump is measured, e.g, the first data point is the difference in fitness between the first substitution and the ancestor.