Research project methodology
The starting point is the UNRES package for large-scale coarse-grained simulations of protein structure and dynamics (available at www.unres.pl). The package has already been extended to include restraints from knowledge-based models of proteins. The Korean partner, which is a leader in template-based modeling of proteins and global optimization, will contribute the recent advancements of the Dynamic Fragment Assembly (DFA) method, the new method to assess the strength of the restraints derived from bioinformatics models, as well as global-optimization techniques, which will be used for force-field optimization, and the global-optimization-based algorithm of handling inconsistent geometry restraints from experimental data. The Korean group will also contribute their recently developed method to study conformational transitions, which is based on a Lorentzian switching function.
The predictive function of the UNRES package will be enhanced by:
optimization of the UNRES force field subject to restraints from bioinformatics-based protein models,
a new algorithm for selection of appropriate consensus fragments to derive restraints from inconsistent bioinformatics-based models,
improvements of the DFA method to derive bioinformatics-based pseudopotentials.
These improvements should result in enhancing the resolutions of the models. The new protocol for the prediction of the structure of proteins will be tested in the Community Wide Experiment on the Critical Assessment of Techniques for Protein Structure Prediction (CASP13 and CASP14) experiments. The developed method for studying conformational transitions will be applied to selected biological problems (transition between the two states of adenylate kinase, between the closed and open form of the Hsp70 chaperone, and between the conformational states of motor proteins).
Expected impact of the research project
The functioning of living organisms is largely dependent on the fact that each of its constituent proteins adopts a unique structure during folding under physiological conditions. Development of an integrated, physics- and bioinformatics-based approach of modeling protein structure and dynamics is of crucial importance to advance our understanding of biological systems. We expect to create a robust package with which to predict the structure and simulate dynamics of very large protein systems such as, e.g., chaperone machinery, mitochondrial respiratory complexes, etc.
International collaboration
Each of the two (Polish and Korean) groups will work on the tasks of the projects which are within the scope of its expertise. The interaction between the two groups has an about 15-year history. Combining the physics- and the bioinformatics-based approaches strengthened by the cooperation of experts from both research groups, the efficiency of which is guaranteed owing to long-lasting mutual contacts and collaboration, provides a big opportunity for developing a reliable and efficient tool for the prediction of protein structures. Another aim of this cooperative project is to tighten the existing cooperation between the Gdańsk and the Korean groups, to extend this collaboration to other groups at KIAS and Faculty of Chemistry, University of Gdańsk, and to establish new connections between the groups of cooperating parties and other scientists from Poland, Korea, and the neighboring countries.