Modeling imatinib-treated chronic myelogenous leukemia: reducing the complexity of agent-based models

Area of Science:

• Computational biology
• Mathematical modeling
• Oncology

Background:

• Agent-based models (ABMs) are valuable for simulating biological systems like chronic myelogenous leukemia (CML).
• Existing ABMs can be computationally intensive, limiting the scale and scope of simulations.
• Imatinib is a targeted therapy for CML, and accurate modeling is crucial for understanding treatment dynamics.

Purpose of the Study:

• To develop a computationally efficient model for imatinib-treated chronic myelogenous leukemia dynamics.
• To replace a complex agent-based model with a system of deterministic difference equations.
• To enable simulations with realistic cell numbers and direct evaluation of steady states.

Main Methods:

• Developed a model using deterministic difference equations to represent agent clusters.
• Discretized the state space to group individual agents.
• Replaced the agent-based model of Roeder et al. with the new deterministic system.

Main Results:

• The new model replicates the averaged behavior of the original agent-based model.
• Achieved significantly reduced computational cost compared to the original model.
• Demonstrated independence of model complexity from the number of agents.

Conclusions:

• The developed deterministic model offers a computationally efficient alternative for simulating CML dynamics.
• This approach simplifies complex agent-based models and facilitates large-scale simulations.
• The reduced computational burden allows for comprehensive sensitivity analyses of model parameters.