Inference after experiments can be difficult. For example, consider the simple observation that reduction in one molecule (x) increases the levels of another molecule (y). 
One inference is that x inhibits y. But, another equivalent inference is that y activates x, which competes with the autoactivation of y via z. Simulations help with such alternative inferences and guide the design of future experiments.
Entity-Sensor-Property Systems: This model simulates entity-sensor-property systems and their evolution across generations, which can be used to explore regulatory architectures. Here are 78,285 sets of parameters that will result in systems that are stable for more than 25 generations. The code can also be downloaded from github and run using Netlogo 6.1.1 offline.
Transgenerational regulation of a germline gene: This model simulates the activity of a gene expressed within the germline across generations in C. elegans, which can be used to discover possible regulatory architectures that are compatible with observed RNA silencing phenomena in this organism. Here are 11 system-ids that will result in parameter sets that support the stability of the system for more than 10 generations. The code can also be downloaded from github and run using Netlogo 6.1.1 offline.