NetLogo User Community Models
by Gary An (Submitted: 07/05/2007)
WHAT IS IT?
This is a simple model of the acute inflammatory response to infection. This model is an extremely reduced version of the "Innate Inflammatory Response" model seen on the Community Models page from May 2004, and contains only some essential components of the body's response to infection. The background patches of the model represent a generic tissue that contains a "tissue-life" patch variable that keeps track of the damage done. There are two breeds, the "inflam-cells" and the "infect-agents." The former represents generic white blood cells that respond to infection/damage and produce a "cidal-compound" that both kills bacteria/infect-agents but also damages normal tissue; this is consistent with the super-oxide radicals and proteolytic enzymes produced by neutrophils in acute inflammation. The "infect-agents" will damage the generic tissue, replicate at a set rate and can be allowed to produce a "toxin" that increases the lethality of the bacteria. The "inflam-cells" will follow a chemotactic gradient based on the "cidal-compound" and thus aggregate in areas of higher activity. In this fashion the "cidal-compound" also represents pro-inflammatory cytokines/mediators that promote the inflammatory response.
This model does not include healing.
Note that this model is extrememly abstract, but does contain the basics of the initial, acute inflammatory response, and as such can potentially be expanded to represent increasing levels of detail in both cellular functions and bacterial properites. This model was used as a teaching tool at the University of Pittsburgh Biodefense Immune Modeling Summer School in June, 2007 to introduce students to agent-based modeling.
HOW TO USE IT
Button "Setup" resets the model.
THINGS TO NOTICE
Note that the initial response to the infection is pretty brisk, and early killing is fairly pronounced. However, also note that there is a concentration of the inflam-cells in areas of highest concentration of the cidal-compound, resulting in a relative depletion of available inflam-cells elsewhere. When the infect-agents on the patches with these high levels of cidal-compound are all killed, you will then see "burst" patterns of release inflam-cells.
Also note that there is a period of dynamic containment where the spread of the infect-agents is stopped, but the central area of infect-agents is persistant. This is analagous to abcess formation.
Note that the infect-agents have a fixed replication rate for the entire population, therefore on the Infect-agent plot you will see intermittant stepped increases in the infect-agent population.
At high levels of infect-agents the model runs very slow due to the overall number of agents. This can be adjusted for in extensions to this model by treating the infect-agents as a patch variable instead of individual agents. However, for purposes of this demonstration it is useful for novice modelers to think of the bacteria as agents in of themselves.
THINGS TO TRY
You can vary the "Initial-infection-number" to find a tipping point between where inflam-cells win or the infect-agents win.
You can add the "toxin" to see the effect of increased lethality seen in bacteria that produce a toxin (such as Anthrax or toxin-producing Staph).
You can also adjust the "Chemotaxis-threshold" to adjust inflam-cell responsiveness (this will be negatively fed-back in realistic situations) or adjust the "Cidal-compund-production" slider to vary the intensity of the inflam-cell response per agent. Notably, increasing the intensity of the inflam-cell response increases the likelyhood of controlling the infection, HOWEVER, increases overall tissue-damage. Thus, even this simple model demonstrates the Janus-faced nature of acute inflammation that leads to clincial conditions such as sepsis and multiple organ failure.
EXTENDING THE MODEL
As mentioned above, this is an introductory toy model of acute inflammation that can be used as a template to model more specific celllular and bacterial responses/characteristics. Additional inflammatory cell species can be added, anti-inflammatory aspects can be added. The sliders of inflam-cell responsiveness have be switched to turtle-variables and varied based on positive and negative feedback loops.
This is a extrememly reduced version of the Innate Inflammatory Response model seen on the Community Model webpage from May 2004.
(back to the NetLogo User Community Models)