3D-FunSites - a virtual encyclopaedia of protein functional sites for assessing variant impacts and drug repurposing

We will establish computational protocols for detecting mutations in gene coding regions which destabilise or modify the structure/function of the translated protein. Such changes can affect interactions of proteins with compounds or other proteins in their cellular environment, leading to disruption of cellular systems and human disease. We will use our CATH resource to detect these pathogenic mutations. CATH classifies evolutionarily related proteins into sub-families (FunFams) and inherits functional properties across relatives e.g. residues involved in catalytic activity, significantly increasing knowledge of key functional sites beyond the 1% experimentally characterised. We will massively expand FunFams and increase functional site information by integrating data from PDBe-KB. By calculating proximity of disease mutations to functional sites we can predict their pathogenicity. FunFams can also be used to repurpose drugs between relatives sharing high structure similarity, to combat mutation impacts. We will take account of gender/ethnic differences to support personalised medicine.