Figure 2
From: Reaction-contingency based bipartite Boolean modelling
![Figure 2](http://media.springernature.com/full/springer-static/image/art%3A10.1186%2F1752-0509-7-58/MediaObjects/12918_2012_Article_1108_Fig2_HTML.jpg)
Comparison of the bipartite and the classical Boolean approaches. A simplified model of the Pheromone response (MAT) and High osmolarity (HOG) pathways, which share two common kinases (Ste20 and Ste11). This toy model consists of two parallel pathways: the MAT pathway is triggered by a pheromone signal ([Pher]) and initiates mating ([MATING]), and the HOG pathway is induced by hyperosmotic stress ([Osmo]) and triggers turgor recovery ([TURGOR]). The network structure of (A) the bipartite Boolean model and (B) a classical Boolean model have topologies corresponding to the regulatory graph (A) and reaction graph (B), respectively, of the same rxncon network definition (Additional file 2: Table S2). The bipartite Boolean code was automatically generated as described in the Methods section, while the classical Boolean code was created manually based on the reaction graph topology. (C, D) State evolution of the MAPK network displayed as heat map. Yellow and blue indicate active (True) and inactive (False) nodes, respectively. (C) The bipartite Boolean model can distinguish different input signals and activate only their specific outputs accordingly. (D) The network structure of a classical approach is simpler, as it does not consider states and reactions as separate from components, but it is not able to maintain signal specificity and always activates both outputs in response to either input signal.