With the advances in analytical techniques and data mining, the chances to elucidate a plant metabolome and understand the metabolite variation in response to external stimuli such as stress are gradually becoming feasible in a global manner. This approach represents a very important research challenge because of the structural diversity of the metabolites and the convolute nature of biological matrices. Based on a new collaborative framework emerging from the recent creation of the EPGL, a project aimed at the development of methodology for the treatment of complex metabolomic datasets with data mining has been initiated with the expertise of different EPGL laboratories. In this paper the strategies used for the study of the metabolic variations in a biological model (the effect of mechanical wounding on Arabidopsis thaliana) are described. Metabolite profiling based on micro-extraction and LC/MS analysis with various detection methods has been used. Data were parsed in the form of ion maps and treatment of this complex set of MS information was performed with dedicated bioinformatic tools. This approach should enable the evaluation of metabolome variations in a comprehensive manner for a better understanding of complex biological mechanisms and for the detection of novel bioactive molecules