Using machine learning to extract models from observations of biological tissues

A fundamental challenge in biology is to understand how molecular networks encode the dynamic behaviors that emerge at subcellular, cellular and multicellular scales. These are critical for regulation of cellular and tissue dynamics that, in turn, drives organismal development, physiological homeostasis or adaptation. Understanding how these dynamics emerge from interactions at microscopic/molecular scales, and encoding this understanding in predictive mathematical models, is the key to predicting and engineering the dynamics of living systems, including designing effective interventions for pathologies. I will discuss data-driven biophysical modeling approaches we are developing to learn the mechanical behavior of adherent cells.