Geometry determination algorithm

The preceeding discussion has briefly described the potential functions and geometry determination algorithms employed in CAMSEQ and particularly in CAMSEQ/M. The "batch" version of CAMSEQ currently in use is totally compatible with CAMSEQ/M. CAMSEQ/M can be interfaced (via a communications link) to a host computer for more lengthy calculations where speed is a requirement (rather than economy). 

Conformational Sampling Conformational Search Proteins Distance Geometry Theory, Algorithms, and Chemical Applications Macromolecular Structures Determined using NMR Data NMR Refinement Simulated Anneeding. 

Around 1985 the first three-dimensional structures of macromolecules determined by NMR and distance geometry were presented (see Distance Geometry Theory, Algorithms, and Chemical Applications). In the meantime, NMR instruments have been greatly improved, new and more sophisticated NMR experiments have been introduced, and various numerical algorithms have been designed for the calculation of structures based on NMR data. Nonetheless, the entire field of NMR. structure determination. 

Experimental methods are available for determining protein structures. X-ray crystallographic analysis provided the first protein structures and remains the most precise method for structure determination. NMR can provide information as to the distances between pairs of atoms in the protein in solution (see Distance Geometry Theory, Algorithms, and Chemical Applications). The information provides restraints on the possible structure of the protein and can be used to obtain refined three-dimensional protein structures. 

An algorithm has been proposed for determining the reaction coordinate, transition structure, and optimized geometry all in a single calculation. The... 

Harris KDM, Johnston RL, Habershon S (2004) Application of Evolutionary Computation in Structure Determination from Diffraction Data 110 55-94 Hartke B (2004) Application of Evolutionary Algorithms to Global Cluster Geometry Optimization 110 33-53... 

The finished network automatically reflects the characteristics of the data domain Not only do the network weights evolve so that they describe the data as fully as possible, but so also does the network geometry. The size of the network is not chosen in advance and as topology is determined by the algorithm and the dataset in combination, it is more likely to be appropriate than the geometry used for a SOM, especially in the hands of an inexperienced user, who might find it difficult to choose an appropriate size of network or suitable values for the adjustable parameters in the SOM. 

---