Data-Driven Analysis of the Trichoplax Adhaerens

Abstract

Biological systems, particularly animals, are a hotspot of highly sophisticated and non-trivial dynamics. They consist of a myriad of motile self-driven and simultaneously interacting entities. Understanding animal collectivity is of central importance for learning evolutionary foundations and inspiring artificially engineered systems. Several models have been developed to describe the dynamics of animal behavior. Recently, advances in data and image acquisition techniques have introduced novel ways of analyzing complex animal groups. In particular, data-driven approaches are increasingly being used to infer partial differential equations (PDEs) for systems of living organisms. In this work, we examine the behavior of a peculiar marine, multicellular animal, the \textit{Trichoplax Adhaerens}. Our objective is to investigate its behavior in two distinct states, one with food present and another without. We use velocity field data generated from two experimentally recorded movies of the \textit{T. Adhaerens}. We use two physically distinct models that integrate conservation laws, and principles of hydrodynamics to infer a set of PDEs modeling the evolution of the datasets at hand. Furthermore, we compare the performance of each model to determine which one best describes the animal's true biology. Finally, we perform modal analysis to identify dominant patterns of motion and derive a set of ordinary differential equations (ODEs).