Dynamics of the laboratory drift.

<p><b>a</b>. Cumulative mutational loads (average number of mutations per gene) along the 17 rounds of the laboratory neutral drift. <i>N</i>_<i>t</i> is the average number of total mutations per gene (shown as ‘diamonds), <i>N</i>_<i>a</i> is the average number of nonsynonymous mutations per gene (shown as ‘squares’). Mutational loads were derived from deep-sequencing of G0, G3, G7 and G17 repertoires (full points) as well as by Sanger sequencing—standard, full-length sequencing of randomly selected variants from each round (empty points). Error bars show the standard error for the calculated averages. The lines illustrate the observed trends (not a fit for a specific equation). <b>b.</b><i>N</i>_<i>a</i><i>/N</i>_<i>s</i> is ratios of nonsynonymous to synonymous mutations (shown as ‘triangles); and the average number of compensatory mutations per gene (<i>W</i>_<i>rel</i> >1.1, shown as ‘circles’). Compensatory mutations are listed in <b><a href="http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1004421#pcbi.1004421.s013" target="_blank">S2 Table</a></b> and were defined as enriched mutations, either by assigned beneficial fitness effect for individual mutations by (<i>W</i>_<i>rel</i> >1.1) or high positional fitness effect (the averaged <i>W</i>_<i>rel</i> per position as calculated in <b><a href="http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1004421#pcbi.1004421.g003" target="_blank">Fig 3A</a></b>, <i>W</i>_{rel (Positional)} >1.1). The effect of compensatory mutations discussed in the section of “<i>Dynamics of the laboratory drift</i>”. <b>c.</b> The cumulative mutational load for mutations with different fitness effects: ‘deleterious’ (<i>W</i>_<i>rel</i> ≤0.6), ‘Nearly-neutral’ (<i>W</i>_<i>rel</i> 0.61–0.8), ‘Neutral’ (<i>W</i>_<i>rel</i> 0.81–1.1) and ‘Beneficial’ (<i>W</i>_<i>rel</i> >1.1).