Research

Mathematical models of how tumors interact, adapt, and respond to treatment.

We study cancer as an ecological and evolutionary process. Our models are developed with experimental and clinical collaborators and are used to formalize biological hypotheses, interpret data, and identify mechanisms of treatment response and resistance.

Research programs

Questions that connect theory, experiments, and treatment

Our work is organized around interactions among heterogeneous cancer populations, their local environments, and the selective pressures imposed by therapy.

01

Spatial microenvironments and residual disease

Tumor behavior is shaped not only by cancer-cell-intrinsic properties but also by interactions with stromal, immune, and other neighboring cells. We study how these local interactions alter competition, selection, and treatment response.

Work in lung and prostate cancer examines spatially restricted protection from therapy, stromal contributions to residual disease, and the relationship between local ecology and the evolution of aggressive phenotypes.

02

The bone ecosystem and treatment resistance

Bone is a dynamic ecosystem in which tumor cells interact with osteoblast-lineage cells, osteoclasts, mesenchymal stromal cells, and treatment. We develop spatial and multiscale models to investigate how these interactions influence tumor establishment, bone remodeling, relapse, and drug resistance.

This program includes multiple myeloma and bone-metastatic prostate cancer and is developed in close collaboration with experimental researchers.

03

Eco-evolutionary dynamics under treatment

Treatment changes the selective environment experienced by a heterogeneous tumor. We investigate how pre-existing variation, environmentally mediated protection, phenotypic state, and acquired resistance jointly shape population dynamics during and after therapy.

These models provide a framework for asking when resistance is likely to emerge and how treatment timing, combinations, or interruptions alter evolutionary outcomes.

Modeling approaches

The method follows the biological question

We use complementary mathematical and computational approaches, selecting the level of abstraction and spatial detail needed for each system.

Evolutionary game theory

Models of frequency-dependent interactions among cancer phenotypes and cells in the tumor microenvironment.

Spatial and agent-based models

Explicit representations of individual cells, local interactions, spatial structure, and heterogeneous treatment exposure.

Dynamical and multiscale models

Mechanistic models that connect processes operating from cellular interactions to population and tissue dynamics.

Model–experiment integration

Iterative work with experimental and clinical collaborators to refine assumptions, test predictions, and interpret observations.

Selected work

Representative publications

View the complete publication record on Google Scholar →