Uncertainty analysis and core predictions

Biological systems are inherently complex systems that pose difficult challenges in the measurement domain, especially in vivo, meaning that often few properties can be measured. Furthermore, most available observations are measured indirectly or under specific experimental conditions, limiting the applicability and quality of the observations. Together, these limitations mean that all parameters in a model might not be uniquely determined from the experimental data. There are different types of analyzes one can perform but generally, they aim to deduct how stable the model structure is and how well defined the parameters of the model are. Well defined parameters are of interest since this is a measure of how well the model replicate the real system. If one would use a developed model a commercial setting, i.e., a sensor or medical product, it is of course important that the model is reliable. Well defined parameters and model behavior is typically need for a model to be of use outside an experimental setting. For a more detailed explanation we refer the reader to one doctoral thesis of a previous PhD in our group, Model-Based Hypothesis Testing in Biomedicine: How Systems Biology Can Drive the Growth of Scientific Knowledge.