Next: density.grid_search() dock Up: The density class: handling Previous: density.grid Density Contents Index
This command reads a density map from file, which should be provided as a grid of intensities, in the X-PLOR [Brünger, 1992], CCP4, or MRC format. (Note that CCP4 and MRC files are currently read in the same way, since the formats are very similar.) file can be a filename or a readable file handle (see modfile.File()).
Note that while non-cubic grids can be read in, density.grid_search() only works with cubic grids.
The size of the grid and the map voxel size (voxel_size) are both taken from the density map file itself. If, however, em_map_size is specified and is non-zero, the grid must be cubic and all dimensions must match this value. If voxel_size is specified and differs from that given in the file header, the user-specified value is used in preference.
The density map resolution is given by resolution.
px, py, and pz specify the origin of the map, in angstroms. If not given, it is read from the MRC or CCP4 map header (X-PLOR files do not include this information, so the origin will default to zero in this case.)
When fitting the probe into the EM grid, the probe structure is converted first into probe density, by using the function indicated in the density_type variable. Each atom can be represented by one of several atomic density functions, including, the uniform sphere model ('SPHERE'), the Gaussian function ('GAUSS'), a normalized Gaussian function ('NORM'), a hybrid Gaussian/sphere model ('HYBRID'), and an interpolation to the closest point on the grid ('TRACE'). The recommended function is 'SPHERE'.
filter_type is used to filter the values of the EM density during this calculation. Filters that can be used are: 'THRESHOLD' for a lower threshold (any density value below the first value set in filter_values will be set to 0); 'SQUARE' for a square filter; and 'LAPLACIAN' for a Laplacian filter. 'NONE' is the default option, and means that no filter is used.
When calculating the cross-correlation coefficient between a probe model and the density map, ccf_func_type specifies if you want the normalized cross-correlation coefficient ('CCF'), or the local cross-correlation coefficient ('LCCF').
Example: See density.grid_search() command.