Perturbation events are a means to change the rate constants of individual rules in the middle of a dynamic simulation.  Each perturbation is applied when a particular trigger condition is met. Perturbations can be triggered at a particular time in the simulation, or the first occasion that an observable increases or decreases past some threshold. 

Perturbations can be added via the "+ Add" button, and can be removed via the (x) next to the perturbation.  Each perturbation has a trigger condition that is specified on the first line, and an assignment specified on the second line.  The drop down field on the trigger condition allows the user to select a trigger based on time, or based on any of the observables specified in the model.  Additionally, the second drop down field lets the user specify whether the perturbation is applied when the trigger is greater than (">") or less than ("<") the values that they enter in the text field. 

The second line of the perturbation allows the user to select a rule from the model whose rate constant should be changed in the event that the trigger condition is met, as well as the new value for the rate constant.  Note that the rate constant must be entered in the units of seconds and molecules.  For example, a first order rate constants would be in units of s^-1, and a second order rate constant in the units of molec^-1*s^-1.

Usage of perturbations

One common use of perturbations is to simulate the addition of a ligand (or inhibitor) at a particular time during a simulation.  There are a couple of ways to do this.  First, the ligand can exist in the initial conditions, but the rate constant for its binding can be set to zero in the rules section of the model page.  The perturbation can then be used to set the binding rate constant to its normal value at the trigger time.  This approach is inconvenient if there are many binding rules for the ligand, and does not work at all if the user desires to use a perturbation to increase the amount of the ligand from one non-zero concentration to another.

The second way to use perturbations to simulate the addition of a ligand is to create a synthesis rule for the ligand that has a rate constant of zero, and use two perturbations to turn on and off the synthesis rule appropriately.  The first perturbation would set the synthesis rate constant to some high rate at the trigger time during the simulation.  The second perturbation would then trigger when the total amount of the ligand (which needs to be tracked via an observable) exceeds the target amount, and would set the rate constant for the synthesis rule to 0.  This method works well for varying the ligand concentration between two non-zero levels, and can equally be applied to ligand removal with the use of a ligand degradation rule. 

It is also important to note that each perturbation is only be applied once during a simulation so, for example, perturbations cannot be used to generate oscillatory behavior in a model of cell cycle progression.