Insight in the functioning of a biological system is gained when it is perturbed and the dynamics of its response are studied. Just like the quality of the analysis of the experimental data, the study design influences the amount and quality of information obtained from an experiment. Optimal experimental design optimizes the information content of a study using a limited amount of experiments. Questions that are answered with optimal experimental design are for example what the optimal frequency and timing of sampling is and what number of individuals and repeats are required. In metabolomics studies the challenge in optimal experimental design lies in the expected multivariate nature of the (unknown) effect of a perturbation and the nonlinear dynamic behaviour of biological systems.
An example of a typical application where experimental design can help greatly, is a challenge test. In such an experiment, subjects are given a dose of a substance (e.g. glucose). After this perturbation samples are taken to monitor the metabolic response and kinetic parameters of absorption and elimination can be estimated. The precision of the parameter estimates depends on the timing of the samples.
Another example where experimental design is of great benefit are batch fermentations where inhibition of growth may occur, but where the precise mechanistic of this inhibition is unknown. With optimal experimental design the sampling times to resolve the inhibition problem can be obtained.