Reconstruction of biological networks through gene silencing and overexpression

Many methods that exist for reconstructing biological networks take into account assumptions of sparsity or already known relationships among states of the network. We show that we can completely characterize the relationships among observed states and inputs in a linearized network through a series of easily performed biological experiments--inducible overexpression and gene silencing. We briefly review dynamical structure and give conditions for a set of experiments that will give all necessary information to obtain the structure.