A fragment library representing available chemicals

The function of many of the large macromolecules that control biological processes can be affected by the binding of small molecules. This is how most medicines work and much of the early stage research in the pharmaceutical industry focuses on identifying and optimising small molecules to be tested as drugs. Over the past ten years, there has also been an increase in what is known as chemical biology research, where the action of such small molecules is used as a tool in fundamental research to understand how biological processes work. However, it is extremely difficult to find a small molecule with exactly the right shape and properties to bind strongly and specifically to a particular macromolecule. One new technique that has been developed recently is the method of fragment-based discovery. Instead of having to find the complete molecule that fits the binding site, this approach begins by identifying smaller pieces of molecule that bind. If the way in which these small fragments can be understood, then the chemist can design changes that merge or grow these fragments into the larger compound with the correct properties.

Fragment-based ligand discovery has recently become validated as an effective method in drug discovery. The central premise is that a small library of low molecular weight compounds can sample a potentially huge chemical diversity. The structure of fragments binding to an active site can then guide the evolution of the fragments into more potent compounds. This provides a tractable alternative to screening for inhibitors, particularly for academic groups that cannot afford access to high throughput screening and for identifying compounds that bind to new classes of protein. One attractive route to evolving fragments is to use the sub-structure of the fragment to identify larger compounds in the available database. There are numerous examples where such larger compounds bind with sufficient affinity to act as inhibitors, allowing the effect of inhibiting the target to be measured in vitro. However, to date, most fragment libraries have been constructed against physico-chemical rules with additional selection on the basis of their tractability for further optimisation by synthetic chemistry. The objective of this grant is to develop and validate methods for constructing a fragment library that is selected to maximally sample the available chemical libraries. We will assess three different cheminformatic approaches to design of a library, select one for implementation, assemble and prepare a library and screen against one target. The resulting hits will then be used to select larger compounds for purchase and testing.