Speaker
Description
Melanoma is a type of skin cancer that becomes much more lethal when it spreads or metastasises to other tissues. During tumour initiation, melanoma cells form clusters in the primary tumour that promote metastasis. In the absence of biological tools to visualise cluster formation at primary tumour sites we use in vitro data for two distinct melanoma cell phenotypes, one more proliferative and the other more invasive. Data is available for both monoculture experiments, which give rise to homogeneous clusters, and co-culture experiments, which result in heterogeneous clusters.
We develop a suite of mathematical models to generate mechanistic insight into the cluster formation observed in the data. We use a coagulation-fragmentation-proliferation framework to describe cluster growth dynamics both in a deterministic mean-field ODE model and a stochastic agent‑based model (ABM) to investigate motility-driven interactions. We quantify phenotypic differences by fitting the ODE models to experimental data using a Bayesian framework for parameter inference and information criteria to perform model selection. To investigate spatial and collision‑driven effects that the ODE framework cannot represent, we analyse the ABM through global sensitivity analysis and derive analytical links between its parameters and the effective coagulation rate in the ODE model. These results provide biological insight into how motility, collisions and proliferation jointly drive cluster formation.
