Not all cancer cells are born equal. It could be demonstrated that tumors are formed by different cell types, mostly exhibiting a limited potential of proliferation. Further, it was shown that malignant tumor behaviour is fuelled by a minority of key-players harbouring the ability to form new tumors after transplantation. Such tumor founding cells share multiple characteristics with stem cells, and thus, are often called “cancer stem cells (CSC)”. With this regards, most anticancer treatments fail to be curative because they focus on the large cell population, but not on CSC. This illustrates the need to develop more efficient therapies. However, such developments should be preceded by (1) better understanding of the mechanisms controlling normal stem cell fate decisions, (2) identification of the mechanisms that differ between CSC and their normal counterparts.In the present project, we plan to characterise mechanisms implicated in self-renewal and differentiation of neural stem cells (NSC) and to compare these data with brain tumor stem cells (BTSC). For this, we aim to establish microarray based transcriptomic and miRNA signatures of NSC, and of cells for which neuronal and astrocytic differentiation has been induced. To these signatures, we want to compare the transcriptomic profiles of BTSC, and thus, to isolate common and differentiating molecular events. The transcriptomic profiling will be completed by studying the impact of specific chemical compounds on stem cell behaviour. In fact, identifying drugs that artificially induce stem cell differentiation might be useful to induce differentiation and exhaustion of the CSC population. Finally, we aim to establish in vitro and in vivo models to verify the functional aspects of the obtained data. Together, our data would have an extensive impact on understanding the mechanisms implicated in NSC fate decisions and in BTSC development, a first step towards the development of efficient therapies.