176
X-Machines for Agent-Based Modeling: FLAME Perspectives
• Function programming: Develop the agent function code.
• Messages between agents: Through the design phase, the input and output messages involved with the agents need to be identified and then
linked with the functions.
• Determine agent memory: For each agent, identify the memory variables.
These will form part of the function code, being manipulated through
messages and agent functions during the simulation.
• What to measure: Identify the model output variables that will be
recorded as simulation objectives. These output variables can either be
average agent behavior on one variable or a number of variables which
change during the simulation to understand agent behavior against iteration time. These outputs can then be compared to real system data
to compare how accurate the model represents the real system.
FLAME has been very successful in modeling a variety of biological experiments. Working with various biologists and involved in their projects, it has
studied systems such as epithelial tissue healing, bacterial concentrations in
oxygen-starved environments, ant and pheromone behavior and even sperm
behavior in reproductive systems. It has aided in unlocking interesting biological phenomena, just by the exercise of conceptualizing and writing models,
with comparing simulated to real collected data. Some of these projects, in
collaboration with experimental biologists, are summarized below.
7.1 Example Models
7.1.1 Molecular Systems Models
Innate immune system. NFκB pathways and its relationship with the cytoskeleton. Nature is governed by local interactions among lower-level
subunits, whether at the cell, organ, organism or colony level. Adaptive
system behavior emerges via these interactions, which integrate the activity of the subunits. To understand the system level it is necessary
to understand the underlying local interactions. Successful models of
local interactions at different levels of biological organisation, including epithelial tissue and ant colonies, have demonstrated the benefits
of such ‘agent-based’ modeling. Here, the modelers presented an agentbased approach to modeling a crucial biological system, the intracellular NFκB signalling pathway. The pathway is vital to immune response
regulation, and is fundamental to basic survival in a range of species.
Alterations in pathway regulation underlie a variety of diseases, including atherosclerosis and arthritis. The modeling of individual molecules,
X-Machines for Agent-Based Modeling: FLAME Perspectives
• Function programming: Develop the agent function code.
• Messages between agents: Through the design phase, the input and output messages involved with the agents need to be identified and then
linked with the functions.
• Determine agent memory: For each agent, identify the memory variables.
These will form part of the function code, being manipulated through
messages and agent functions during the simulation.
• What to measure: Identify the model output variables that will be
recorded as simulation objectives. These output variables can either be
average agent behavior on one variable or a number of variables which
change during the simulation to understand agent behavior against iteration time. These outputs can then be compared to real system data
to compare how accurate the model represents the real system.
FLAME has been very successful in modeling a variety of biological experiments. Working with various biologists and involved in their projects, it has
studied systems such as epithelial tissue healing, bacterial concentrations in
oxygen-starved environments, ant and pheromone behavior and even sperm
behavior in reproductive systems. It has aided in unlocking interesting biological phenomena, just by the exercise of conceptualizing and writing models,
with comparing simulated to real collected data. Some of these projects, in
collaboration with experimental biologists, are summarized below.
7.1 Example Models
7.1.1 Molecular Systems Models
Innate immune system. NFκB pathways and its relationship with the cytoskeleton. Nature is governed by local interactions among lower-level
subunits, whether at the cell, organ, organism or colony level. Adaptive
system behavior emerges via these interactions, which integrate the activity of the subunits. To understand the system level it is necessary
to understand the underlying local interactions. Successful models of
local interactions at different levels of biological organisation, including epithelial tissue and ant colonies, have demonstrated the benefits
of such ‘agent-based’ modeling. Here, the modelers presented an agentbased approach to modeling a crucial biological system, the intracellular NFκB signalling pathway. The pathway is vital to immune response
regulation, and is fundamental to basic survival in a range of species.
Alterations in pathway regulation underlie a variety of diseases, including atherosclerosis and arthritis. The modeling of individual molecules,