Supplementary Figure 11: Xanthophore movement models with our Box initial condition. We present (a)–(d)
EA ABM results based on 5000 simulations, (e)–(h) the corresponding PDE solution under the optimal value
of αXX for a mesh resolution of Nbin = 240 voxels, (i) the distribution of nearest cell–cell distances for an
example ABM simulation at 150 days, and (j) the PDE solution with αXX = 1 at 150 days for comparison. A
difference in cell density of 772 cells/mm2 corresponds to about 0.12 cells per voxel for this choice of mesh
resolution. In (a)–(h) and (j), we overlay cells positions (in red) for an example ABM simulation; we remove
the upper region of cells in post-processing for clarity. As we show in (h) and (j), the support of our PDE
solution with optimal parameters is about 69 μm closer to the ABM support (at the example region shown); this
is a difference of about 2 xanthophores based on empirical estimates [20].
EA ABM results based on 5000 simulations, (e)–(h) the corresponding PDE solution under the optimal value
of αXX for a mesh resolution of Nbin = 240 voxels, (i) the distribution of nearest cell–cell distances for an
example ABM simulation at 150 days, and (j) the PDE solution with αXX = 1 at 150 days for comparison. A
difference in cell density of 772 cells/mm2 corresponds to about 0.12 cells per voxel for this choice of mesh
resolution. In (a)–(h) and (j), we overlay cells positions (in red) for an example ABM simulation; we remove
the upper region of cells in post-processing for clarity. As we show in (h) and (j), the support of our PDE
solution with optimal parameters is about 69 μm closer to the ABM support (at the example region shown); this
is a difference of about 2 xanthophores based on empirical estimates [20].