Computational modelling is a powerful approach that complements experimental investigation to derive understanding of biological systems. We have developed a methodology based on automated formal reasoning, which enables the synthesis of biological networks that are consistent with current experimental data, and which can be used to predict untested behaviour. In collaboration with Austin Smith’s lab at the University of Cambridge, we previously applied this approach to study the gene regulatory networks governing self-renewal. This talk will present our methodology in the context of new investigations, where we have sought to expose the logic of network resetting for induction of naïve pluripotency. We have found that a Boolean network architecture defined for self-renewal accurately predicts reprogramming potency of single or combinations of factors. Deterministic gene activation trajectories were computationally identified and experimentally substantiated at single cell resolution. We tested 132 predictions formulated by the computational models, finding a predictive accuracy of 79.6%, and further show that this network accurately explains experimental observations of somatic cell reprogramming. We conclude that a common and deterministic program of gene regulation governs both self-renewal and induction of pluripotency.