Analytic formulation of optimal structural intervention strategies in Boolean Networks

Boolean Networks (BNs) as simplified gene regulatory network models enable us to study their dynamics, which is critical for developing future genetic therapeutic intervention. In this paper, we alter the long term behavior of BNs by modifying the underlying regulatory rules, which we term as function perturbations. As the long-run behavior of BNs reflects the phenotype of the cell, we search for the optimal way to perturb BNs to avoid undesirable aberrant cellular states, such as cancer. For a network constructed from melanoma gene expression data, the optimal function perturbation has been found based on the derived results to achieve a more favorable steady-state distribution and reduce the risk of metastasis in melanoma.