In lack of quantitative data ad-hoc parameters were used. For different initial conditions

numerical simulations of the model were run until convergence. Each initial condition was determined by eight values — the levels of
Otx2 and
Gbx2 on the anterior as well as posterior side of the boundary and the levels of
Fgf8,
Wnt1,
En and
Pax, whose initial expression domains were all set to the four central compartments. Hence, the initial conditions were not biased towards the desired expression pattern of these genes. (A) Classes of steady states reached by the model. Dark grey: expression, light grey: no expression but diffusion of secreted protein, white: neither expression nor diffusion. (B) On the left-hand side a scatter plot of the reached steady states is shown. The X- and Y-axes mark the difference between the initial concentrations of
Otx2 and
Gbx2 in the anterior and posterior side of the neural plate/tube. The Z-axis marks the minimum of the initial concentrations of
Fgf8,
Wnt1,
En and
Pax on a log-scale. Each marker in the scatter plot corresponds to one of the

model simulations. Its spatial position is determined by the initial conditions of the model variables, its color by the steady state that was reached, cf. A. For better visualization the cube on the left-hand side was sliced up horizontally (middle column). The distribution of the six steady states from A in each slice is shown on the right-hand side. Details about the ad-hoc parameters, the sampling of the initial conditions and the classification of the steady states can be found in Text S1.