Multi-agent based modeling of liver detoxification

Using game theory and reinforcement learning, we created and analyzed generalized agent-based models of hepatic toxin elimination processes to explore plausible causes of hepatic functional zonation. We considered a general situation in which a group of protective agents (analogous to liver cells) cooperate and self-organize their efforts to minimize optimally the negative effects of toxin intrusions. The model suggests that in order to do so, the agents should adjust their resource consumption based on two factors: 1) their ranked proximity to the common wealth and 2) the potential damage caused by toxins. We verified that liver cells do the same.