Identifying, predicting and controlling ecological systems that generate spatio-temporal point patterns

Recent technological advances have resulted in a multitude of spatio-temporal cell imaging data. These can be translated into spatio-temporal point patterns in which points represent cells. Such data hold rich information about how cells act and interact. However, much of this information is not interpretable by current methods.

This data-driven observation motivates our main objective, which is to develop methods that (1) probabilistically identify ecological systems from the spatio-temporal point patterns that they generate and (2) enable prediction and control of identified systems.

We will use two unified classes of mathematical models to describe ecological systems: spatio-temporal point processes (STPPs) and spatial cumulant models (SCMs).

STPPs are governed by probabilistic rules that describe how points appear and disappear on spatio-temporal point patterns.

SCMs are formulated by partial differential equations that approximate the dynamics of first and second order spatial cumulants, i.e. densities and spatial covariances. The STPP-SCM unification enables tractable mathematical treatment of STPPs, which we will leverage for system control.

We will test our identification, prediction and control methods on experimental cancer cell ecologies, via integrated theory-experiment approaches. This will directly evaluate the applicability of our methods.

Sara Hamis (Uppsala University) and a PhD student will work on the project in collaboration with multidisciplinary experts.