The effects of mutations are modeled as a reflected gamma distribution. Use the sliders to set the values of the parameters:
Choose the single-mutation distribution tab to see the effect distribution for the chosen parameter values. Choose the multiple-mutation distribution tab to see simulated combined effects of multiple random (non-neutral) mutations per strain.
The plots show the selection coefficient, $s$, which is defined such that $g_{MA} = g_{anc} \left( 1 + s \right)$, where $g_{MA}$ is the growth rate of a mutation-accumulation strain (or a hypothetical strain with one non-neutral mutation) and $g_{anc}$ is the growth rate of its ancestral strain. The reflected gamma distribution does not stop at values less than $-1$, which are biologically unrealistic because they imply a negative growth rate. Therefore, for the purposes of the multiple-mutation simulation, any simulated $s < - 1$ is set to $s = -1$ (lethal). This truncation is rare but is most likely to happen when $k$ is small and $m$ is large. The reflected gamma distribution also allows large positive values. Such values are biologically implausible but are nonetheless possible in principle, so the positive values are not truncated in the simulation. The combined effect of multiple mutations is the product of the $\left( 1 + s \right)$ values for each mutation, minus 1 (there are no interaction effects).