Three Glazier-Graner-Hogeweg (GGH) simulations wrapped as process-bigraph Processes using CompuCell3D. Each configuration demonstrates a distinct multicellular behaviour with interactive 2-D lattice visualization.
Two cell types self-organise via Steinberg sorting
Two cell populations (TypeA and TypeB) are initialised as a mixed blob on an 80x80 pixel lattice. Differential contact energies cause TypeA cells (low homo-adhesion, J=2) to engulf TypeB cells (higher homo-adhesion, J=16), reproducing the Steinberg differential adhesion hypothesis. After ~2000 MCS the cells segregate into distinct clusters.
Cells chase a secreted chemical signal
TypeA cells (green) secrete a diffusible factor "Signal" that TypeB cells (orange) follow via chemotaxis (lambda_chemo = 500). The diffusion solver runs a forward-Euler PDE on the same lattice. Over time, TypeB cells cluster around the TypeA secretors, demonstrating gradient-driven cell migration on a 2-D CPM lattice.
Population expansion via cell mitosis
Cells start from a small blob and grow by incrementing their target volume each MCS (+0.1 px/MCS). When a cell's actual volume exceeds 50 pixels it divides randomly, producing two daughter cells that reset to the base target volume. The colony expands outward as the population doubles repeatedly, demonstrating tissue growth on a 150x150 lattice.
Growing colony with chemotaxis-driven dispersal
This experiment combines all three physics modules — differential adhesion, chemotaxis, and mitosis — to model tumour spheroid invasion. TypeA cells (green) form a tightly cohesive core (J_AA = 2) that secretes a diffusible signal. TypeB cells (orange) are repelled by the signal (lambda_chemo = -200) and have weak medium adhesion (J_M-B = 8), causing them to scatter outward as invasive cells. Both populations grow and divide (division at 50 px, +0.06 px/MCS). The result is a dense expanding core surrounded by a diffuse front of dispersing invaders — a hallmark of collective cell invasion.