phenix_logo
Python-based Hierarchical ENvironment for Integrated Xtallography
Documentation Home
 

Creating maps using the Phenix GUI

The "Create Maps" GUI
Guidelines for use

Because of the differences in calculation method and input data, there is no single unified interface for creating any electron density maps; instead, separate programs are grouped together in the main GUI. For omit maps, a modified interface to AutoBuild is available; consult the AutoBuild documentation for rules on use. Isomorphous difference maps have a separate GUI, which takes two reflection files as input. For maps using phases from a model, use phenix.create_maps. All of these programs write output as an MTZ file; the latter two may optionally output XPLOR maps (calculated to cover the PDB file used for phasing) if desired.

The "Create Maps" GUI

The interface for calculating electron density maps uses the same code base as the validation tools and consequently has a similar interface. The primary difference is the ability to request arbitary map types, including user-defined difference maps. The window shown below has the default maps entered: 2mFo-DFc and mFo-DFc difference maps.

images/map_types.png
The pre-defined maps include all common likelihood-weighted difference maps, anomalous difference map, and likelihood-weighted F(obs) or F(calc) maps. Custom map types can be defined by typing in an appropriate string in a format similar to those already shown; the map generator will parse the string to determine the components used. If the "Save maps in XPLOR format" box is checked, maps suitable for display in PyMOL will be generated. You may define custom maps, using the same map type syntax as for MTZ format, but with additional control over the region covered by the map and the grid scale. If no additional XPLOR maps are requested, Phenix will simply convert any map coefficients from the MTZ file to XPLOR format (covering the input molecule plus a buffer region).

Guidelines for use

  • "Kicked" maps use small random perturbations in coordinates to reduce model bias and noise; the final map is an average of independently calculated perturbed maps. This adds to calculation time, but is considerably faster than a composite omit map.
  • Map quality may be heavily influenced by reflections that are missing due to errors in data collection or processing. Substituting F(calc) for missing F(obs) increases model bias but often results in a more easily interpreted map; the default maps output by phenix.refine (and REFMAC) are "filled" maps. In phenix.create_maps, the default maps are filled, and we recommend using this option unless you are particularly worried about model bias.
  • Output MTZ column labels must be unique; the predefined map types have associated predefined labels, but you can change these if you wish.
  • Anomalous difference maps require anomalous data (i.e. F+,F- or I+,I-); the precalculated anomalous difference DANO generated by some CCP4 programs is not suitable.
  • If you are using neutron data, set the scattering table to "neutron"; otherwise, leave it set to "n_gaussian".