"The activity of many proteins, strongly changes when small ligands dock into well defined cavities of the protein receptor. For proteins involved in disease-related metabolic pathways novel artificial ligands may act as drugs. In the drug discovery process, high-throughput in-silico screening(IS-HTS) methods are increasingly used to identify lead compounds, i.e. ligands that bind to a target protein with high affinity. Existing methods have significant shortcomings, because the time spent on each ligand is very short. In an ongoing collaboration with the Barcelona Supercomputer Center (BSC) I implemented FlexScreen, the IS-HTS program developed by the host group at FZK, on the Cell, a new processor developed by BSC and IBM into the next generation of ultra-fast peta-flop supercomputers. The individual processing unit of these new computers is distributed separately and delivers about 250 times the performance of a PC at roughly the same cost. In the mobility project I will exploit this unprecedented computational power to incorporate new ideas into the IS-HTS method, not achievable with standard computational hardware. Continuing the collaboration with BSC I will develop improved biophysical scoring functions and novel docking strategies. I will also address a long-standing problem of IS-HTS methods: the change of protein conformation upon binding (induced fit). In collaboration with the cancer research center in Heidelberg (DKFZ) I will apply these new methods to search for new lead compounds to DNA-methyl-transferases to develop new therapies for cancer. The results of this research will reduce the time and cost of the initial stages of drug design, not only in academic but also for industrial projects, increasing the competitiveness of the European pharmaceutical industry. They will improve the quality of life of the European population by helping to discover new drug molecular scaffolds, addressing one of the main problems of present-day drug development."