The Thomson Lab is applying quantitative experimental and modeling approaches to gain programmatic control over cellular differentiation. He is developing mathematical models to ask how cellular regulatory networks generate the vast diversity of cell-types that exists in the human body. He is applying models to engineer and rewire cellular physiology and to synthesize new types of cells that do not exist in nature. He is also developing simplified cellular systems in which physical models can be applied to control the geometry and morphology of different cell types. He uses a combination of approaches including mathematical modeling, machine learning, statistical analysis of high-throughput gene expression data, and single cell RNA sequencing experiments. Recent accomplishments include: Engineering an all-optical differentiation system in which he could optically-deliver pulsed neural differentiation inputs to embryonic stem cells; creating new computational tools for deriving cell state trajectories from single cell RNA-Seq data; and developing a stochastic modeling framework for analyzing principles that enable robust self-organization of the mammary gland.