PROTEIN STRUCTURES: THEIR VALIDATION FOLD CLASSIFICATION AND INTERACTIONS. Janet M. Thornton1,2, E.G. Hutchinson1, S. Jones1, R.A. Laskowski1,2, M. W. MacArthur1, A. Michie1 and C.M. Orengo1, 1Department of Biochemistry and Molecular Biology, University College, Gower Street, London WC1E 6BT, 2 Department of Crystallography, Birkbeck College, Malet Street, London WC1E 7HX
As the number of known protein structures rises rapidly, we begin to appreciate the extent of the universe of protein folds. For each new structure apart from its unique biological interest, it is essential to validate the co-ordinates, to describe the structure in terms of its relationship to other known structures and their functions, and to study the interactions made by the protein with other biomolecules. Since structures are often solved very rapidly it has become essential to develop approaches and software tools which facilitate automated validation, analysis and classification of the structure for the crystallographer. These tools are valuable for users of the databank .
In this presentation, three different aspects of structural analysis will be considered. First, methods for validating structures, based on current knowledge derived from known structures will be discussed. These methods complement the critical measure of agreement between the model and X-ray data. Recent results derived from very high resolution structures will be presented. Secondly, as the size of the database grows it becomes more difficult to know whether a structure is novel or has been seen before. Our hierarchical description of protein structure in terms of Class, Architecture, topology and homologous superfamily (CATH) will be described. Lastly, novel approaches to identifying active sites and recognition patches on the surface of a protein will be discussed.
Information available at http://www.biochem.ucl.ac.uk/bsm