The immediate goal of this project is to develop a physics/chemistry-aware inverse reinforcement learning (IRL) computational framework to discover how heterogeneous cancer cells function singly or collectively to drive cancer progression. The long-term goal of this research centers on understanding single-cell and cooperative decision-making that drive tumor growth, metastasis, and recurrence. The proposed work is computational science in nature, developing new and scalable artificial intelligence (AI) algorithms that leverage cell imaging data to extract knowledge on cancer cell behavior and predict interventions.
The project team hypothesizes that heterogeneity in cancer arises through predictable processes, producing a cohort of team players that function collectively to drive critical steps in disease progression (survival, proliferation, invasion, and metastasis), measuring “success” on a tumor-wide rather than individual cell basis. The team also hypothesizes that the motivations behind cell decisions, represented as a reward function within a physics/chemistry-abiding multi-agent model, can be learned via IRL from real-world spatial-temporal image data containing thousands of live cells uniquely available to the team.