On the Mechanics of Cellular and Multicellular Active Matter
Name
yang-hqyang-phd-meche-thesis-Feb2026.pdf
Size
37.64 MB
Format
Adobe PDF
Checksum (MD5)
b5bb8dbbae9280eb42155ef7fba75bf9
Author(s)
Yang, Haiqian
Advisor(s)
Guo, Ming
Date Issued
February 2026
Publisher
Massachusetts Institute of Technology
Abstract
How does biological life gain its form and shape? Watching an embryo develop under the microscope is like watching a 4-D video playing in real time. Individual cells sense and respond to their environment, and together they form robust structures at the tissue scale. Today, we still lack a predictive framework to model their behaviors, either at single-cell or multicellular scale. In the first part of the thesis, we focus on a mechanistic view. In close conjunction with experiments, we derive continuum-mechanical theories to understand the mechanics of cells and tissues, from the visco-hyperelastic cytoskeletal networks, to the nonlinear cell-matrix interface, towards the activity-induced spontaneous orientational ordering in cell-matrix mixture. These works provide a mechanistic understanding of the complex physical responses of these active-matter systems. In the second part of the thesis, we primarily focus on confluent tissues, ranging from cell monolayers to Drosophila embryos. An important element of this part is abstracting living tissues as graphs. We first examine the Delaunay tessellation of tissues and propose the volume and shear order parameters, and we further investigate the probability distribution, leading us to the configurational entropy and temperature, using which we quantify a wide range of living tissues in silico, in vitro, and in vivo. These ultimately lead us to building graph-based data-driven methods to study the dynamics of tissues. To this end, we introduce graph neural networks to study the glassy dynamics of cultured cell monolayers, reaching good performance, and we further build MultiCell, a model that predicts different types of cell behaviors at single-cell resolution and over time during the dynamic process of Drosophila whole-embryo development. Together, these works set the stage for predictive modeling of multicellular active matter.
MIT Department
Massachusetts Institute of Technology. Department of Mechanical Engineering
Terms of Use
In Copyright - Educational Use Permitted
Copyright retained by author(s)
Persistent DSpace Link
