SEF Spencer, SM Hill and S Mukerjee
Dynamic Bayesian networks for interventional data
Abstract: Graphical models are widely used to study biological networks. Interventions on network nodes often form part of experimental designs for the study of biological networks. In this paper, we discuss the modelling of interventional time-course data using graphical models known as dynamic Bayesian networks. Our work is motivated by, and illustrated in the context of, protein signalling networks. We show empirical results that demonstrate the need to carefully model interventions when interventions form part of the design.