Towards modularity in biological networks while avoiding retroactivity

We approach the solution to the problem “are biological networks modular” using a systems theory approach. We propose a method to divide a particular family of gene regulatory networks into “modules” that are functionally isolated from each other, so that the behavior of a composite network of two or more modules can be predicted from the input-output characteristics of the individual modules. This method provides a platform for the creation of new foundational modules using which networks can be decomposed. We present our work for the deterministic case, while providing a well-known example from the literature to validate our approach.