Mark Smythe

Combinatorial chemistry

Current strategies in library design involve the calculation of hundreds of potential descriptors that define various chemical characteristics, and selecting a diverse set of compounds in this descriptor space. With hundreds of available descriptors it is difficult to know which descriptors, if any, are important or essential for describing biological activity. Consequently such procedures result in the optimisation of libraries in chemical descriptor space, which has little impact on biologically relevant regions of that space. To overcome this we have developed a series of biologically-relevant descriptors that are used in library design. As a consequence, we aim to identify the biologically-relevant structural regions of chemical diversity and design and synthesise arrays of molecules that match this diversity space. We have developed new linkers and auxiliaries to aid combinatorial synthesis and a molecular design platform to achieve these objectives. We have synthesised various constrained cyclic peptide libraries (molecular toolkit libraries) and libraries of macrocycles and heterocycles. We have a particular focus on the discovery and exploitation of privileged structures.