Structural Bioinformatics

Manfred J. Sippl

Center of Applied Molecular Engineering, University of Salzburg

The protein structure initiative (PSI) is an international effort to determine as many protein structures as possible by X-ray analysis and nuclear magnetic resonance. The ultimate goal of PSI is to make feasible the prediction of accurate three-dimensional structures of proteins based on their sequences alone. This will require a data base of structures with at least one high-resolution experimental structure for each protein family - defined as a group of proteins sharing more than 30% sequence similarity. Currently PDB, the protein data bank, contains the structures of 25,000 proteins. These structures are the result of more than fourty years of protein structure determination. These and the new structures determined over the coming years need to be classified, analysed, and used for the prediction of new and unknown protein structures and their functions. The major computational techniques required in these efforts are structure comparison, structure search, error detection and correction, and structure modelling and prediction. These research areas are main subjects of 'Structural Bioinformatics'. Over the last decade the Center of Applied Molecular Engineering at the University of Salzburg contributed several key technologies to structural bioinformatics, like knowledge based potentials derived from protein structures, error detection in experimentally determined protein structures, analysis of protein stability, structure comparison, and classification of protein structures. In this talk I review these topics and discuss current developments.

Email: sippl@came.sbg.ac.at

Address: Jakob Haringer Str.5, A-5020 Salzburg,