The central class in the Admin package is ExperimentSet, whose instances are used to represent a homogenous set of scientific experiments. These experiments are performed under the same conditions, which are formally described using the ExperimentProtocol class and each experimental protocol follows a certain Method. Grouping such experiments into sets removes the need to specify these conditions repetitively for each individual experiment. Also, a single Person is in charge of controlling these experiments within a single Lab. Individual experiments, represented by instances of the Experiment class, differ only by their chronological features and in relation to their subjects. Sets of experiments of different types are grouped into an ExperimentalStudy, which summarises a specific metabolomics study. The model does not explicitly differentiate between different types of experiments (i.e. biological sample preparation, chemical sample preparation, analytical experiment and data analysis) at the class level. In other words, a single class is used to describe all four types of experiments, because there are no structural differences at this level. The differences arise only in relation to other classes. For example, an analytical experiment is performed on an analytical Instrument, while data analysis is performed using a Program. Since the same instrument is used to perform a set of analytical experiments, there is a link between the ExperimentSet and Instrument classes. This is in contrast to a link between a single Experiment and the biological sample (the BioSample class in the Biological sample package) it produces.