Sampling Properties of Simulated Annealing and Distance Geometry
Properties of two different methods for calculating three-dimensional structures of macromolecules from nuclear magnetic resonance data, distance geometry, and simulated annealing, are studied. It is shown that a simulated annealing refinement of structures generated with distance geometry is sufficient to remove bias introduced in the distance geometry stage. A new efficient simulated annealing protocol which does not require initial structures from distance geometry is presented. The influence of distance selection on quality and distribution of structures generated with distance geometry is studied.
KeywordsSimulated Annealing Nuclear Magnetic Resonance Data Ramachandran Plot Distance Restraint Distance Geometry
Unable to display preview. Download preview PDF.