Figs. S1 to S6 Tables S1 to S4 References

Mutation accumulation Eight independent lines were derived from a single haploid ancestor derived from FY10 (Mat-a, leu2Δ1, ura3-52) (1), a strain isogenic with the reference strain s288c. These lines were maintained on standard YPD (yeast extract-peptone-dextrose) solid plates. Approximately every 20 generations, a single colony was used to produce new colonies in the next generation by streaking and plating cells. This propagation scheme represents a single-cell bottleneck every 20 generations, resulting in an effective population size of ~10 as calculated using the harmonic mean of the population sizes. At this population size, the fixation or loss of most mutations will be determined by random genetic drift and not natural selection. Only selection coefficients larger than the reciprocal of the population size (10%) have a significant effect on the fate of new mutations, and these are expected to be rare (2, 3). There is no direct estimate of the rate of nucleotide substitution in yeast, but the per-genome mutation rate has been estimated to be around 0.003 in S. cerevisiae (4). Pairs of strains that have diverged for 8,000 generations are therefore expected to differ by ~30 nucleotide substitutions. Other perturbations are als