Control-compatible state reduction for Boolean networks

Reduction of gene regulatory networks is a key problem for genomic signal processing, in particular, to overcome the computational limitations to the design of optimal intervention policies for large networks. This paper proposes a greedy procedure to reduce the state space of a Boolean network with perturbation based on the basins of attraction in the network. We study the effects of the reduction on the steady-state mass for attractors and on the intervention performance of a greedy stationary control policy that is also dependent on the basins of attraction.