Fig. 4: From reactions to a functional pathway.

We used the smoothed update rules to generate and simulate a model of the Sln1 branch of the high osmolarity glycerol (HOG) pathway. a The pathway visualised as a rxncon regulatory graph. In the absence of turgor, Sln1 stays unphosphorylated. As turgor increases, the auto-phosphorylation of Sln1 initiates a phosphotransfer cascade converging at Ssk1. The phosphorylated form of Ssk1 turns off the downstream MAP kinase pathway leading to dephosphorylation of the downstream transcription factor Hot1. The dashed line indicates a feedback loop that is included in the cyclic version only. The model has 29 different targets: 12 reaction targets (pink nodes), 16 state targets corresponding to 15 elemental states (blue nodes) and a component state for the phosphatase (not displayed in the graph), and the single input/output target “Turgor” (grey node). b Simulation of a linear version of the model using source state smoothing of the update rules. (I) We use our default assumptions on the initial state and simulate the model until we reach an attractor (first OFF trajectory). (II) We activate the system by turning [Turgor] ON and simulate again (ON trajectory) until we reach an attractor state. (III) From there, we set [Turgor] OFF again and simulate the model until we reach an attractor. We observe that the model responds as expected to the input. c We extended the HOG model with a feedback loop, where activation of the pathway leads to increased turgor (via Hot1-{P}). This is simplification of an adaptive response through increased glycerol production and retention, which increases turgor. We simulate this model from the initial OFF attractor (see panel B), and note that the system oscillates as expected: the trajectory is cyclic, where the last time step is followed by the first, and turgor is now a model variable that turns on and off during the cycle (grey row in the heatmap).