Well developed structured models can tell us a great deal about the dynamics of viral infection within cells. Models of this nature can lead to better antiviral drugs by revealing where the weakest links are in viral synthesis. Unfortunately, many difficulties can arise when creating a structured model, including obtaining rate constants and concentration data for each step and finally validating the mechanism of infection. We have developed a structured stochastic simulation to test some of these issues. Influenza A was chosen since its infection mechanism is relatively well documented and there is large amount of published data on its various rate constants; although there are many values that remain ambiguous. Using various statistical experiments and response surface models of all the constants in the simulation, over 45 in total, a “best guess” at the remaining constants was made. Under the assumption that nature tends toward optimizing its processes to edge out competition, the optimization of the response surface models given various constraints should result in calculated rate constants that closely mimic the 'true' constants. This methodology can be applied to the mechanism itself. It stands to reason that if the proposed mechanism used in the simulation cannot be solved with given real world constraints then an alternate mechanism should be proposed. Additionally, experimentally determined rate constants can also be used to validate the optimized results.

Extended Abstract Status: Not Uploaded