MAST: a hybrid Multi-Agent Spatio-Temporal model of tumor microenvironment informed using a data-driven approach

Area of Science:

• Computational biology
• Cancer research
• Systems biology

Background:

• Computational modeling, including agent-based models, is used to study tumor-immune cell interactions in cancer.
• Each tumor has a unique microenvironment, necessitating personalized study approaches.

Purpose of the Study:

• To present MAST, a hybrid Multi-Agent Spatio-Temporal model for simulating unique tumor subtypes and tumor-immune dynamics.
• To enable data-driven simulation of cancer scenarios using high-throughput sequencing data.

Main Methods:

• MAST combines a discrete agent-based model with continuous partial differential equations.
• The model captures essential tumor microenvironment components.
• It is informed by high-throughput sequencing data for personalized simulations.

Main Results:

• MAST was applied to human colorectal cancer data, simulating four distinct subtypes.
• The model investigated spatio-temporal evolution and emergent properties.
• Results align with existing tumor biology knowledge and clinical outcome data.

Conclusions:

• The MAST model demonstrates validity for simulating personalized tumor-immune dynamics.
• It offers a valuable tool for understanding unique cancer microenvironments.
• The approach supports data-driven insights into cancer subtypes and clinical outcomes.