January 1, 2010
December 31, 2012
€116,000 (2,919,000 CZK)
We have selected the 3D Voronoi diagrams as the powerful data structure sufficient for solving many protein-scan problems including the above introduced channel search. Currently, in our prototype implementation we are able to compute exactly all channels leading through given protein molecule to an active site in a very short time period. Moreover, the Voronoi diagrams are also suitable for solving the Collision-Detection (CD) problem, which can shift the horizons in protein investigation.
For protein engineering, the very important problem is to determine the narrowest part of the channel during the time. This information can be used either to increase the radius of the channel or to close the channel. This is achieved by two-step process. First, it is necessary to determine the residues of the protein around the narrowest place; then a chemist would try to replace them by the more suitable residues by selecting their rotamers (different positions of the residue in the space defined by the rotation angles).
New methods will be integrated into a tool called CAVER (developed at MU), which is currently used by biochemists around the world. This project is a continuation of the project GA201/07/0927 Visualization of protein structures from years 2007-2009.