During natural senescence of Arabidopsis thaliana, a number of phenotypical changes can be seen as the plant tries to reabsorb nutrients from ageing leaves. Very little, however, is known about the underlying genetic changes which are occuring and even less about the interactions between these genes. A large quantity of highly replicated microarray timecourse data in a complex design has been collected from leaves of Arabidopsis which will make it possible to quantify these genetic changes. By developing MAANOVA, a package for The R Project, it is possible to remove the complexity of this microarray data and obtain a normalised data set over the timecourse which will be indicative of genetic changes due to biological causes alone; removing experimental influence and errors. This data will then be processed using Variational Bayes State Space Modelling to identify likely gene interactions between genes showing most significant expression changes over the timecourse. The result will be a network map of the gene interactions within which some genes will be nodes; influenced by or influencing a large number of genes within the network. By using knockout mutants and over expressors, these node interactions can be confirmed or dismissed in order to generate a more comprehensive priors list and refine the network.