-
what environmental inputs are required to evolve coupled interactions, rather than a colony of autonomous oscillators?
- evolve under normal light/dark, then with different noise for each cell, with cost function requiring that the cells are on according to the mean dawn/dusk
- grid of cells with identical networks
- to start with, for a specific interaction (say A -> B), then for each interaction (A -> B):
- add an additional parameter that determines whether its origin node (A) effects the 'B' genes in neighbouring networks
- to start with, use majority rule as is coded currently (i.e. more activation than repression => activation and vice versa), then if too simple, can set threshold for how much external/internal A is necessary for switching B on/off,
- then boolean decision for B according to inputs
-
can we evolve quorum sensing - evolving the ability to sense when you have many neighbours
- if there are only a few then gene A stays off
- if there are many then gene A turns on