Pbn-Stac: Deep Reinforcement Learning-Based Framework for Cellular Reprogramming

Cellular reprogramming can be used for both the prevention and cure of complex diseases. However, the efficiency of discovering reprogramming strategies with classical wet-lab experiments is hindered by lengthy time commitments and high costs. In this study, we leverage deep reinforcement learning to develop a novel computational framework that facilitates the identification of reprogramming strategies. To this end, we formulate a control problem in the context of cellular reprogramming for the Boolean and probabilistic Boolean network models of gene regulatory networks under the asynchronous update mode. Furthermore, to facilitate scalability, we introduce the notion of a pseudo-attractor and a procedure for the identification of pseudo-attractor states. Finally, we devise a computational framework for solving the control problem, which we test on a number of biological networks.