Experimental modelling is the process of mapping the various steps of an actual experiment onto the structure supported by the database. In database terminology, the set of tables and the permitted connections between them is called the schema. The schema used by maxdLoad2 is designed to allow a wide variety of experimental configurations to be represented. Whilst this is clearly desirable, it does mean that there are often several plausible ways of modelling a given experiment and that identifying the 'best' choice can be hard.
The portions of the schema which cover the microarray decription and the hybridisation process are unlikely to pose any significant modelling problem as they will be the same for any experiment. It is the remainder of the schema, the parts which used to capture the various phases of treatment, splitting and pooling of the biological assets, that will require the most modelling effort.
Each component of your experiment that you wish to record should be matched to one of the tables in the schema. The components are both tangible assets such as specific organisms and pieces of hardware, and the processes and protocols that were applied to them.
A useful starting point is to draw a diagram of your experimental process. The diagram should illustrate the 'workflow' that will be undertaken; starting from the acquisition of source organisms or materials, proceeding to the various treatments that are applied to differentiate the samples (for example exposure to a stressor) and finishing at the point were genomic material is prepared for hybridisation.
This diagram can now be used as a starting point for constructing the mapping from your materials and processes to the various concepts that the maxdLoad2 schema provides.