Results of robustness analysis for hypothesis (1) using the reward operator.

<p>Hypothesis (1) requires stabilisation of in the low concentration mode (). A CSL formula with cumulative reward operator is used in this case. Each of the curves represents the evaluation function over degradation obtained for a particular setting of . More precisely, the horizontal axis shows the perturbation of <i>pRB</i> degradation rate and the vertical axis shows the probability of the hypothesis to be satisfied. In the upper left corner, robustness values are shown for each of the curves. The values are displayed with the absolute error quantifying the precision of the approximate method. For comparison, the values are computed also on piece-wise affine approximations of the evaluation function. It can be seen that the robustness values change rapidly with different settings of . This observation goes with the fact that with faster degradation of there is a higher probability that the positively self-regulated protein is locked in the stable mode of no production. The decrease of the value with increasing is due to the weakening effect of inhibition by <i>pRB</i>.