What about robustness ?
In this part, the idea is to estimate the range allowed to make a parameter varying. Thanks to the criteria used, we can draw a little map of suitable parameters. As a result, we can obtain this kind of pictures:
We will then run simulations for all the parameters together. Obviously, the results are not going to be easy to show. Therefore, clustering algorithms will be used, in order to present to the user just one example of each clusters found. So, we will be able to identify broad sets of usable parameters.
The last part consists in deciding which sets are biologically relevant (knowing some parameters for instance). This is likely to provide interesting estimations of unknown values.
| 
"
