Databases and Grid computing are a good match. With the service orientation of Grid computing, the complexity of maintaining and integrating databases can be kept away from the actual users. Data access and integration is performed via services, which also allow to employ an access control. While it is our perception that many proposed Grid applications rely on a centralized and static infrastructure, Peer-to-Peer (P2P) technologies might help to dynamically scale and enhance Grid applications. The focus does not lie on publicly available P2P networks here, but on the self-organizing capabilities of P2P networks in general. A P2P overlay could, e.g., be used to improve the distribution of queries in a data Grid. For studying the combination of these three technologies, Grid computing, databases, and P2P, in this paper, we use an existing application from the life sciences, drug target validation, as an example. In its current form, this system has several drawbacks. We believe that the...