Towards Quantitative and Predictive Models of Tumour Ecology: A Framework for Calibrating Evolutionary Game Theory with Experimental Data

MaximilianStrobl

Cleveland Clinic

"Towards Quantitative and Predictive Models of Tumour Ecology: A Framework for Calibrating Evolutionary Game Theory with Experimental Data"

Tumours are complex ecosystems where diverse cancer cell subpopulations interact with each other and with non-cancer cells around them. Evolutionary game theory (EGT) has established itself as a powerful mathematical framework to study the implications of such ecological interactions, demonstrating an important role in shaping oncogenesis and treatment response. However, much of this work has been theoretical using parameters that are only loosely grounded in biological data. To move towards quantitative and predictive models of tumour ecology it is crucial to develop theoretical and experimental methodology to empirically calibrate and validate EGT models. We present an in silico study to optimize the 'Game Assay' for measuring ecological interactions between cancer cell populations in vitro. This assay, originally developed by Kaznatcheev et al (2017), involves co-culturing populations at different ratios, monitoring growth rates via time-lapse microscopy, and inferring frequency-dependent interactions. We begin by characterizing the accuracy and precision of this assay in a simulation study in which we use the replicator equation as the “ground truth”. Our simulations reveal potential biases in estimating fitness differences and interaction parameters, highlighting the need for careful experimental design. We provide guidelines for optimizing seeding ratios, number of replicates, and frequency of measurements, and present a new analysis techniques to improve the accuracy and precision of interaction measurements. Finally, we apply our optimized protocol to quantify interactions between 4 drug-sensitive and resistant lung cancer cell lines, revealing diverse ecological dynamics. This work demonstrates the power of integrating mathematical modeling with experimental approaches to develop robust empirical protocols and gain a quantitative understanding of tumour ecology.
Additional authors:

---