University of Dundee

Global analysis of drug binding sites

We are a multi-disciplinary laboratory where students have the opportunity to develop creative methods to improve drug discovery by applying techniques from chemoinformatics, structural bioinformatics and machine learning. In addition the Hopkins laboratory house the largest high content phenotypic screening and biophysical screening centres in European academia.  


Polypharmacology is a common trait of drugs and offers both opportunities for therapeutic enhancement and challenges of off-target promiscuity. Hence understanding the basis of polypharmacology is essential for understanding the molecular basis of efficacy and safety. However polypharmacology is to be expected given the limited size of the metabolome – the complement of endogenous small molecule ligands – relative to the size of the proteome. Moreover recent studies by Schneider and Skolnick have suggested that there may only be a limited repertoire of small molecule binding sites in proteins; again providing theoretical underpinnings for the phenomena of polypharmacology. Therefore, we postulate there may be a limited and enumerable set of pharmacophores and thus chemical scaffolds to present those pharmacophores for each metabolite or canonical binding site. The project will involve mining ChEMBL and the PDB to link metabolites with binding sites and orthosteric medicinal chemical ligands.  From this information a set of pharmacophores will be created to define canonical binding sites. Similarly, a set of pharmacophores will also be created to represent allosteric binding sites. These will be identified by mining the PDB for ligand binding pockets for which no enzymatic function or endogenous metabolite ligands are known. Once these aims have been achieved, the nature of orthosteric and allosteric pockets and ligands can be compared and contrasted. This information will aid the design of compound and fragment screening collections covering bioactive space as completely as possible. To this end, particular interest will be placed in how canonical pharmacophores can be partially represented by fragments and by larger compounds.

The project offers sabbatical opportunities to undertake research at the pharmaceutical company UCB.

PhD applicant requirements: A background in bioinformatics would be advantageous. Previous experience with chemistry is desirable but not essential. The PhD applicants should possess a desire to apply their computational skills to pharmaceutical research.