Agents and Biological Entities
In an agent-based model of a biological system, agents represent biological entities. What counts as an entity is defined relative to the model itself. In a given model, an entity is any component of the model that can be treated as an individual having a fate independent from the fate of other components of the model. The first step in devising an agent-based model is to identify the entities that can be modelled by agents.
Not all biological entities need to be modelled by agents. We can distinguish between entities that are passive and those that are active.
-
Passive entities have no behaviour of their own; rather, any changes in their state are brought about by other components of the model (e.g. other entities, gloal changes). These can be modelled by objects .
-
Active entities are those that have both state and a defined behaviour . These can be modelled by agents . Agents can act on their environment, which might include objects and other agents.
Agent and object types are specified in the model, each with their own set of characteristics (state variables and state transition rules). During simulation, instances of these agent types are created (the number of instances is specified in the model).
Agent Environments
Agents are both sensitive to and able to modify their environment. Rules govern their interactions with their environment and determine which parts of the system make up their environment. An environment or neighbourhood is defined relative to an agent and can be dynamic i.e. what counts as its environment at one time may not be accessible to it at another time. Neighbourhoods can be defined spatially e.g. nearest neighbour or non-spatially e.g. participation in a signalling network.
The environment of an agent can include:
Next: Mediation of Interactions |
