Screening techniques are routinely applied in drug discovery to search large libraries of drug-sized molecules for suitable lead candidates. While first hits with micromolar activity can often be obtained, considerable difficulties arise upon their optimization due to constraints of size and ADME properties. As an alternative, new approaches to screen collections of small fragments have been developed with the goal to probe a large chemical space with a minimal number of compounds. Fragments detected as hits are then used to assemble larger ligands with improved potency. Due to the small size of the fragments, "drug-like" characteristics are more easily evolved at the same time. The Workshop is planned to discuss the actual impact of current fragment-based techniques on the drug discovery process.
Structure-based design and virtual screening are still mostly applied on enzymes with well-defined binding pockets. However, more and more targets are becoming available which are less straightforward to work with. On the one hand, many structures can nowadays be obtained by increasingly reliable homology modelling, but the value of these models for drug design is still questionable. On the other hand, new structures of complicated systems such as protein-protein interfaces are experimentally elucidated, but not yet easily addressed as drug targets. The Workshop will try to assess current strategies for facing these challenges.
The Workshop is intended to bring together about 90 experts in drug design in an informal atmosphere in order to discuss recent problems and approaches related to fragment-based drug discovery, homology modelling, and challenging targets such as protein-protein complexes.
Fragment-based techniques have gained increasing popularity over the past years. Once only computationally feasible, molecules much smaller than typical leads can nowadays also experimentally be probed for binding to a given target. To profit most from these approaches, theory and experiment need to go hand in hand. What are the chances and limits of computational and experimental methods for fragment-based drug design and how are their forces best joined?
The discovery of new protein sequences and their functional annotation proceeds much faster than the elucidation of the corresponding 3D-structures. However, more and more sequentially related proteins are structurally determined and the chances for successful homology modelling augment. Can we predict protein structures accurately enough to apply the arsenal of computational drug-discovery tools to these models?
The "drugability" of enzymes is well-understood, but nature provides us with many other interesting systems that could be valuable drug targets, such as cell-surface structures, flexible protein domains, and protein-protein interfaces. Can we readily transfer to such systems the strategies developed for enzyme binding pockets or do we need entirely new concepts?
The workshop is organized by: