Fig 1.
Process flow and component diagram for a modeling and simulation workflow with StochSS.
The biochemical model and domain is defined as part of the problem specification. The ODE, spatial stochastic, and well-mixed simulation tools generate realizations of these models. The parameter estimation and parameter sensitivity tools allow for analysis of models. The output and visualization tools present the data.
Fig 2.
Screenshots of the StochSS model editor.
Choose the model to edit from the selection list and view and edit your domain with the mesh editor (left), and define the biochemical species, initial conditions, parameters, and reactions (right).
Fig 3.
StochSS provides built-in visualization capabilities in order to quickly explore simulation results such as deterministic ODE simulations (A) and well-mixed discrete stochastic realizations (B). Using external plotting libraries, in this case matplotlib in Python, we highlight the key qualitative differences between the deterministic and stochastic simulations (C). As can be seen, while the mean values differ slightly between the model levels, the most dramatic difference is apparent when considering individual realizations, which reveals a high noise expression level of transcription factor. For spatial stochastic modeling, the model editor provides the capability to visualize the computational mesh and the subdomains as wireframes (see Fig 2), and simulations can be visualized and animated, in this case using volume rendering (D) and solid rendering with domain clipping (E).
Table 1.
Illustration of simulation times and data output sizes for the different modeling levels supported by StochSS.
There is a steep increase in computational cost as the model is refined.