Combinatorial libraries biased/targeted

A list of privileged scaffolds - several of which are natural-product derived - for target-family-biased combinatorial libraries was recently presented by Muller [94], These scaffolds were proven to produce biologically active compounds for more than one member of a given target family. A rough-and-ready in-silico evaluation... 

Many proprietary and commercially available libraries are synthesized in this combinatorial fashion and so suffer from these limitations. Another criticism of these libraries is that they may be too heavily biased toward certain predefined criteria such as the Lipinski rule of 5. ° These rules for bioavailability have proved very useful in drug discovery however, they may be unnecessarily limiting when it comes to the discovery of ligands for novel biological targets such as protein-protein interactions. Despite these limitations, many of the principles of combinatorial chemistry are shared by DOS. 

---