Abstract:

Developmental Graph Cellular Automata (DGCA) are a type of Graph CA which can modify the structure of the graph through their update rule. This can lead to the phenomenon of graphs dividing into disconnected components. We use this as an endogenously driven analogy of cell division. Each connected graph component (or “cell”) can exhibit distinct dynamics, including entering an attractor cycle. This can be used as an analogy of cell differentiation, in which each graph component can end up with a completely different structure. A single DGCA rule (“genome”) can give rise to a system of dividing graph components (“cells”) each of which may end up in a different attractor with different graph structure to become a fully differentiated cell. All this can arise from a relatively compact growth rule being applied repeatedly to a simple seed graph, showing the emergence of complexity from simple processes that is typical of living systems. We present the results of some experiments to evolve growth rules to produce a desired level of cell differentiation.

@inproceedings(Waldegrave++:2025-ALife,
  author = "Riversdale Waldegrave and Susan Stepney and Martin Trefzer",
  title = "A Model of Cell Division and Differentiation with Developmental Graph Cellular Automata",
  pages = 561-569,
  crossref = "ALife-2025"
)

@proceedings(ALife-2025,
  title = "ALife 2025, Kyoto, Japan",
  booktitle = "ALife 2025, Kyoto, Japank",
  publisher = "MIT Press",
  year = 2025
)