Lipinski and others, through concepts such as drug-likeness, re-focussed drug discovery back to the principles of medicinal chemistry in the high-throughput era as key to reducing attrition. More recently, the need to go further in defining what makes a good lead has been recognised with the concept of leadlikeness. Leadlikeness implies cut-off values in the physico-chemical profile of chemical libraries such that they have reduced complexity (e.g. MW below <400) and other more restricted properties. We examine these concepts in the context of Virtual (theoretically possible), Tangible (chemically feasible) and Real (physically available) worlds of molecules. In a thought experiment, we take the HTS concept to the extreme: screening an estimated 60 million 'Global Collection' on 5000 targets and realising that perhaps millions of drug candidates might be found that could not possibly be handled in reality. Sampling of the Virtual and Tangible worlds is therefore a necessity. We show that the world of Reals is significantly under-sampled as the MW of compounds increases. This supports the design and screening of 'reduced complexity' (leadlike) compound libraries, preferably with synthetic handles available for rapid chemical iteration and detected as interesting by careful screening or biophysical assays.

Bringing new medicines to the market depends on the rapid discovery of new and effective drugs, often initiated through the biological testing of many thousands of compounds in high-throughput screening (HTS). Mixing compounds together into pools for screening is one way to accelerate this process and reduce costs. This paper contains both theoretical and experimental data which suggest that careful selection of compounds to be pooled together is necessary in order to reduce the risk of reactivity between compounds within the pools.

No abstract

No abstract

This paper explores the ability of the 2D-3D structure conversion packages Concord, Cobra, ChemDBS-3D, and Converter to generate the structures of bound small molecules and compares their conformations with those in ligand complex crystal entries in the Brookhaven Protein Data Bank. ChemDBS-3D is limited by the size of structure that it can handle in its database environment and can only process 62% of the structures, but when these structures are compared with the ligand structures, they are found to most closely approximate the bound conformation. However Converter can be considered to perform better as it can convert 100% of the structures from input 2D diagrams, and its conformations are a good match in most cases. Concord performs well in converting 92% of the structures from input SMILES strings, and the conformations that it produces are good in many cases. The version of Cobra used (1.