-
Each cell (that is, patch) is either normal or abnormal. You could have each patch keep track of this with some variable, or you could use color.
-
In each time step, each normal cell divides with some probability birthprob, which is defined on a slider.
-
In each time step, each abnormal cell divides with some probability kappa * birthprob, where kappa is defined on a slider and is greater than 1.
-
Cell division works like this: if a cell divides, choose a random neighbor (use ask random-one-of neighbors [ ]) to change.
-
Have abnormal cells die occasionally. Do this either by making death a random process using a deathprob defined on a slider, or have all abnormal cells die after they reach some age. If you choose age-dependent death, you'll have to have each patch keep track of its age. When abnormal cells die, either have them turn into normal cells, or define a third state: unoccupied.
-
Make border patches blue, or do something else to differentiate the borders. We want to eliminate the wrap-around behavior by making patch behavior dependent on not being on the edges.
-
Many tumor models only examine cells that lie on the exterior of the tumor. That is, only abnormal cells bordering on normal cells can divide, and only normal cells bordering on abnormal cells can divide. (The interior cells divide, but their behavior isn't believed to affect the results. These cells just replace like cells that have died.)
Questions:
-
Under what conditions do tumors grow?
-
Does the behavior change if you only look at exterior cells?
-
Does cell death affect behavior?
-
If cell death does affect behavior, how is the behavior different between probabilistic growth and age-dependent growth? What about probabilistic where the probability depends on age?
