Simulations are defined by their initial conditions, a protocol and an optional rate variant. Use the New Simulation Setup panel to provide these inputs, assign a name to the simulation and determine how any previously defined parameters distribution might be sampled. The Sobol Sequence is typically a good choice for generating samples. For large numbers of parameters the Sobol and Halton sequences suffer from correlations and other dependencies between parameters if the number of samples is small (the exact number depends on the sequense and dimension of the parameter space). The Halton sequence shows the strongest dependence, The Leaped Halton sequence reduces the effect by skipping samples. The Sobol Sequence performs somewhat better than the Leaped Halton sequence. Both the Random and Latin Hypercube option do not suffer from these problems, but both have other drawbacks. The Latin Hypercube is efficient at distributing samples, but once a setup is created no new samples can be added. The Random drawing of samples inherently suffers from the uneven distribution of samples in parameter space, this leads to the well-known one over square root of n behaviour for the convergence of estimates, whereas the other method for generating samples approach one over n.

The defaults in the advanced settings typically don't need to be changed. The Maximum Complex Size limits the size of complexes that will be created by Simmune, this sets an upper bound on the size of the generated network. The Maximum Simulation Duration sets the maximum time the computer is allowed to spend on a single simulation. This prevents spending large amounts of computer time on simulations that accidentally turn out to be much more complex than intended. The relative Tolerance sets the ratio between the error estimate and the value of concentrations at any given time. The integration procedure will adapt its internal step length in an attempt to minimize the computation time while keeping the error below the threshold implied by the relative tolerance. The absolute Tolerance serves a lower bound, below which relative errors that are larger than the relative tolerance are ignored. The Don't generate results for hidden complexes check box is useful for reducing the number of observables that are saved. If the box is checked, no complexes that are marked as hidden in the SimModeler will be written to the simulation output. If the Detect steady state is checked, the simulator will assume that the biochemistry in a simulation is at equilibrium, if the norm of the right hand side is less than 10 % of the relative tolerance. The screen-shot above shows an example simulation using a previously defined initial condition named New Initial Condition using the protocol Stimulus and wash. Only initial conditions defined for the currently active model can be selected using the combo-box. Pressing the Create Simulation Setup button will save the simulation setup in the database and create simulations according to the chosen number of samples and sampling method. The list of defined Simulations can be used to review the settings for previously defined setups. A preview of the simulation results can be viewed if a simulation is selected, the preview is evaluated at the median values of the parameter distributions. The Run Preview button can be clicked to force immediate evaluation the preview simulation on the client workstation, this is useful to reduce the waiting time if a large number of other previews is pending on the server. After the simulation has finished, the Select Items to View box is populated with the observables generated from the Model.

There are three categories of observables. Items listed Complex with molecule represent the concentration of all complexes, that contain the listed molecule, combined. Items under Molecule represent the concentration of the listed molecule either free or in any complex. Reported concentrations will only differ from Complex with molecule when oligomerization is possible. The last option Pattern allows you to access groups of complexes that are described by complexes in the Modeler.

You can select multiple concentrations simultaneously using Ctrl+left mouse button or Shift+left mouse button.