FIGURE 13.

Test stability under wrong assumption of σ_xy. Data sets X_binding were simulated at D_A = 1 μm²/s, D_AB = 0.1 μm²/s, τ_off = 100 ms, τ_on = 200 ms, and t_del = 25 ms and σ_xy = 50 nm. We included 2500 trajectories of exponentially distributed length with mean 10. We tested against the assumption of σ_xy = 25 nm (A), σ_xy = 50 nm (B), and σ_xy = 75 nm (C). Apparently, the test would tolerate such significantly wrong assumptions, yielding extended regions with p-value > α for all settings of σ_xy; yet the significance regions are somewhat biased to lower K and τ_off or higher K and τ_off for underestimation or overestimation of the localization errors. It should be noted, however, that even such significantly wrong assumptions lead only to a minor bias of the results.