Explore the Simulation Menu

The Simulation drop-down menu lets users verify and run simulations.

Click 'Verify' to check the current workspace. A confirmation dialog will appear if verification is successful.

If there are verification errors, the Errors window will open with details.

Click a warning hyperlink to navigate to the warning-causing entity on the canvas (the corresponding Property Inspector will also be presented on the screen).

Select 'Debug Model' to open the Model Debugger modal.

The Model Debugger enables running simulations with enhanced inspection and control compared to normal simulations.

Note: The Model Code Debugger is available only with a license that includes debugging capabilities.

The Neutral Randomisation and Tolerance Scanning windows provide tools to investigate model numerical stability and result robustness, along with relevant evidence.

Select 'Neutral Randomisation' to open the Neutral Randomisation window.

For Neutral Randomisation, specify repeats and enable State Order and/or Rate Contribution Order. State Order randomizes states in the ODE solver array, affecting numerical errors. Rate Contribution Order randomizes contributions to each state's rate, causing different floating point errors.

The plot types available for Neutral Randomisation are:

Tolerance Scanning options include enabling absolute or relative tolerance scanning (one must be enabled) and defining the range for each. Range options include step type (base 10 logarithmic, natural logarithmic, or linear), minimum, maximum, and number of steps. Default ranges use base 10 logarithmic steps with three steps from a tenth to ten times the workspace tolerance.

Available plot types:

Values for one quantity are plotted over time. The Mean with Standard Deviation plot includes error bars.

Both Neutral Randomisation and Tolerance Scanning can produce CSV files with unprocessed simulation results.

Select 'Solver Settings' to view the settings applied to standalone QSP simulations.

The Solver Settings window validates input values. Invalid entries trigger a red border around the text box and an error symbol. Hovering over the error symbol displays a tooltip explaining the issue. Edits cannot be applied with invalid values but can be canceled.

Click the 'Solver Errors' option to display the 'Solver Errors' window, which lists the indices of the individuals for which errors occurred with all the corresponding error flags during the simulation.

Select 'Verification Options' to set the workspace's 'Unit Exponent Tolerance' and 'Frequency Sum Tolerance. ' The 'Verification Options' window will appear.

This window allows setting the workspace's 'Unit Exponent Tolerance' for comparing unit exponents to determine dimensional equivalence. Zero tolerance can be entered for exact equality checks during verification.

Set the 'Frequency Sum Tolerance' to ensure allele frequencies for a genotype sum to one during verification.

Choose whether to show ignorable warnings during verification (unrecoverable errors are always shown).

Select 'Sample Population Model' from the drop-down menu to open the window for saving a sample population file. Enter the number of individuals for the sample model. Clicking 'Sample' opens the 'Save As' pop-up screen to save the Workspace file. Click 'Save' to save the file, or modify the File name field first and then click 'Save. '

Select 'Simulate' from the drop-down menu to open a menu with four simulation options:

Selecting 'Run With Options' from the 'Simulation' drop-down menu opens the 'Run Simulation' window, where different sampling times for various quantities can be specified.

Selecting 'Run' from the 'Simulation' drop-down menu initiates the simulation.

Selecting 'Run VPop' from the 'Simulation' drop-down menu opens the 'VPop Management' window, providing greater control over VPop simulation.

Selecting 'Batch' from the 'Simulation' drop-down menu opens the 'Batch' window to create a batch run with a series of batch items.

The individual fitting functionality in the QSP application is called 'Parameter Estimation' (PE). Select 'Parameter Estimation' to view the Parameter Estimation pop-up screen.

VPop Inference analyzes simulations to determine if certain quantities fall within specified bounds at given time points. This analysis can be run on multiple individuals, categorizing them based on whether they meet the criteria. Select 'VPop Inference' to view the pop-up screen.

Select 'Sensitivity' to view the Sensitivity Window pop-up screen.

QSP Designer enables the creation of complex biological system models, which can be challenging to understand. Sensitivity analysis helps by measuring if model outputs are sensitive to specific inputs. This can determine if certain parameter values affect outputs with available experimental data.

Various methods exist for sensitivity analysis. Simple approaches consider partial derivatives of outputs to inputs, providing local information. The global approaches implemented here sample the entire parameter space within specified ranges, calculating overall sensitivity measures based on numerous simulations.

Select 'Parameter Scanning' to view the Parameter Scan window.

Parameter scanning enables automatic variation of specific parameters and execution of multiple simulations to analyze the model’s response to input changes. Parameters can be chosen from the workspace and sets of values specified either manually or via an automatically determined range. QSP Designer then runs simulations using all possible combinations of these parameters, allowing selected outputs from the model to be viewed.