TY - JOUR T1 - Modeling signaling-dependent pluripotent cell states with boolean logic can predict cell fate transitions JF - bioRxiv DO - 10.1101/115683 SP - 115683 AU - Ayako Yachie-Kinoshita AU - Kento Onishi AU - Joel Ostblom AU - Eszter Posfai AU - Janet Rossant AU - Peter W. Zandstra Y1 - 2017/01/01 UR - http://biorxiv.org/content/early/2017/03/10/115683.1.abstract N2 - Pluripotent stem cells (PSCs) exist in multiple stable states, each with specific cellular properties and molecular signatures. The process by which pluripotency is either maintained or destabilized to initiate specific developmental programs is poorly understood. We have developed a model to predict stabilized PSC gene regulatory network (GRN) states in response to combinations of input signals. While previous attempts to model PSC fate have been limited to static cell compositions, our approach enables simulations of dynamic heterogeneity by combining an Asynchronous Boolean Simulation (ABS) strategy with simulated single cell fate transitions using Strongly Connected Components (SCCs). This computational framework was applied to a reverse-engineered and curated core GRN for mouse embryonic stem cells (mESCs) to simulate responses to LIF, Wnt/β-catenin, FGF/ERK, BMP4, and Activin A/Nodal pathway activation. For these input signals, our simulations exhibit strong predictive power for gene expression patterns, cell population composition, and nodes controlling cell fate transitions. The model predictions extend into early PSC differentiation, demonstrating, for example, that a Cdx2-high/Oct4-low state can be efficiently and robustly generated from mESCs residing in a naïve and signal-receptive state sustained by combinations of signaling activators and inhibitors.One Sentence Summary Predictive control of pluripotent stem cell fate transitions ER -