A compositional framework for Boolean networks

Developing Attractor Analysis Techniques for a Compositional Boolean Network Framework

“…In practice, the application of Boolean network techniques is limited by the well-known state space explosion problem where the exponential growth of the state space severely limits the size of the model. This limitation has led to the development of compositional techniques to support Boolean networks (for example, see [3], [16], [30]). In recent work [3], [4] an interesting new compositional approach was proposed based on composing two Boolean networks by merging entities using Boolean operators such as conjunction.…”

“…This limitation has led to the development of compositional techniques to support Boolean networks (for example, see [3], [16], [30]). In recent work [3], [4] an interesting new compositional approach was proposed based on composing two Boolean networks by merging entities using Boolean operators such as conjunction. This compositional framework set out to provide a foundation for both engineering Boolean networks and decomposing them to aid analysis.…”

“…This compositional framework set out to provide a foundation for both engineering Boolean networks and decomposing them to aid analysis. The initial work so far has focused on the preservation of the behaviour of subnetworks in a composition [3]. However, no support has so far been developed for compositionally identifying the attractors in a composed model.…”

“…In this paper we present the first initial work on developing techniques and tools for compositionally identifying attractors in the given compositional framework [3], [4]. The approach we develop is based on analysing each subnetwork's interference state graph [3], a state graph that extends a Boolean network normal behaviour with the additional dynamics arising from interference in a composition.…”

“…In this paper we present the first initial work on developing techniques and tools for compositionally identifying attractors in the given compositional framework [3], [4]. The approach we develop is based on analysing each subnetwork's interference state graph [3], a state graph that extends a Boolean network normal behaviour with the additional dynamics arising from interference in a composition. We use the strongly connected components (SCCs) [15] in the interference state graphs to identify cyclic behaviour which is then selectively merged based on a new property called interference alignment.…”