Spatial-Temporal Pattern Synthesis in a Network of Locally Interacting Cells

Spatial-temporal patterning is an essential process indicating the cell and tissue types formed during synthetic and natural tissue development. Due to biological complexity within and among cells, most approaches to engineer cell cultures are limited in their ability to control the timing of spatial patterning. This paper presents a computational framework to control spatial-temporal pattern formation in a network of locally interacting agents. Given a parameterized model of patterning in stem cell colonies, we use a supervised machine learning algorithm combined with temporal logics to quantitatively describe and characterize spatial-temporal patterns. We utilize an optimization procedure to achieve optimal parameters that maximize the occurrence of desired spatial patterns at a specified time interval and demonstrate the ability of our algorithm to control transitioning in a sequence of patterns.