Assignment of terminal penalties in controlling genetic regulatory networks

Probabilistic Boolean networks have been used to model inter-gene relationships and their dynamic behavior. Optimal control of such networks has been proposed by turning on or off individual genes. Though this problem was solved in an earlier work by using a dynamic programming algorithm, issues regarding the assignment of terminal penalties and selection of genes for intervention were not addressed. In this work, we provide an algorithm for assigning terminal penalties, taking long term uncontrolled behavior into account. We also discuss the possibility of using gene influence for pre selection of genes to be used for intervention. This is implemented for the popular WNT5A network.