Optimizing and focusing search and results ranking in P2P networks becomes more and more important with the increasing size of these networks. Even though a few approaches have already started to investigate the computation of PageRank-like values in P2P environments, none so far has investigated how personalization could be added to it. This paper tackles the problem of distributedly computing Personalized PageRank values in such a distributed environment and presents an algorithm which uses them to optimize and focus search in the P2P network. The paper also discusses how these algorithms improve current distributed search in power law networks and gives some simulation results.