A new phenotypically constrained control policy in boolean networks based on basins of attraction

In this paper, a new stationary control policy in Boolean networks (BNs) is derived based on the structural properties of basins of attraction (BOA). The new greedy control policy guarantees that no new attractor cycles will be created by intervention to the original BNs. Hence, it will not introduce significant steady-state mass to ambiguous network states, which may represent unknown phenotypes corresponding to potential collateral damage. Comparing the control performance of this new phenotypically constrained BOA control policy with previous control policies based on long-run network behavior, the experiments based on randomly generated networks have shown that the new control policy obtains competitive performance with reduced complexity.