Molecular recognition: automated reconstruction and analyses of large molecular complexes

Consortium leader: Professor MARK JOHNSON
Department of Bichemistry and Pharmacy, Åbo Akademi University

Other project leaders of the consortium: 
Tiina Salminen, Department of Bichemistry and Pharmacy, Åbo Akademi University
Tommi Nyrönen, CSC - Scientific Computing
Cheng Holland, professor, Karolinska Institute, Sweden and University of Calfornia Davis, USA

Doctoral students of the consortium: 
Susanna Repo
Mikko Vainio
Mikko Huhtala
Lenita Viitanen

Key words:  Molecular complexes, interactions, proteins, structure, bioinformatics, cryo-EM, crystallography, modeling

Description of project and main results:
One of the big challenges facing us today is reconstructing larger molecular complexes when x-ray structures are not available for the complex as a whole. Often x-ray or NMR structures are availalble for the component proteins of a complex, and often we have a good idea of the surface regions most likely to be involved in binding, we may also have cryo-electron microscopy data giving a low-resolution view of the complex. Like with small molecule -  protein docking, conformational flexibility of the two molecules makes the automated docking problem a very difficult one.  With protein-protein docking the problem is magnified by the large surface areas involved, the lack of conformational constraints on the interacting side chains, and the likelihood that larger scale conformational changes may take place in one or more of the proteins involved on binding. A variety of methods have been used to tackle the problem, but with marginal success, reflecting our fundamental lack of understanding of the physico-chemical determinants of binding and the effects on local and global conformation. This project will address this problem from several different directions, using both mathematical and computational approaches and experimental methods (x-ray crystallography and cryo-EM).

Publications: