CELL CAD

Changing cell or tissue structures requires changing the molecular state of cells. However, we do not know how to translate cell and tissue structures into the molecular programs that encode them. We propose to take two complementary approaches towards solving this problem: we will build mathematical models from first principles and identify the key parameters in these models that are amenable to programming; and we will train machine learning algorithms to recognize non-linear correlations between molecular state and physical structure that can be exploited for design.