A cellular automata approach to the modelling of cell-cell interactions (Q4377797)

Cellular automata representations of biological cells are typically modelled as one automaton per cell and their interactions are between the nearest and maybe next nearest neighbours. This has several disadvantages, for example: the cells have a fixed size and shape; the dynamics can be influenced by grid effects; all cells interact with a fixed number of neighbours; time and space are highly discretised. Not all of the above points will be important or affect a model's dynamics but it is possible that they could play a significant role in the observed behaviour. A recent paper introduces what is called a ``cell programming language''. A ``cell'', in this language, is represented by a collection of connected squares. The cells can perform global (to the cell) functions, for example grow, divide, roundup, etc. However, there is an inelegant aspect of this model: Before a cell can perform any function it must recompute its neighbourhood to take account of any changes that have taken place since it was last updated. A more elegant model is that of Glazier and Graner (1993) who use subcellular processes. They only considered cell sorting through differential cellular adhesion. We show that their model can be easily and elegantly extended to include such things as growth, division, etc. also on the subcellular level. It then becomes a useful tool in modelling cell-cell interactions and possible self-organized complex behaviours. For the entire collection see [Zbl 0877.00017].