How does it fit into my existing project?



Often, the current data collection is part of a larger project and additional information is available. This could either help the processing of the dataset one loooks at - or it can ensure that the results from data processing are consistent with other data within that project.

When looking at reflection data, there are three main areas one has to be concerned with.

Correct cell and spacegroup for the given crystal form

If a spacegroup allows enantiomorphs (like P41212 versus P43212) - which are indistinguishable at the reflectin file level - the user needs to provide the spacegroup that autoPROC should process the data in. A similar situation can arise, if the current crystal was oriented in a way that didn't allow collection of data along all possible screw axes. This would lead autoPROC assume a spacegroup e.g. of P2 (and not P21), not because the screw axes isn't present in the data, but only because it wasn't observed.

There are two possibilities available in autoPROC to specify cell and/or spacegroup:

  • from the command-line
      % process symm=P43212 cell="77.4 77.4 37.7 90 90 90" ...
    • one could also only give the spacegroup - and autoPROC will select the cell of the best indexing solution given that spacegroup
  • using a reference MTZ file
      % process -ref apo.mtz ...
    • autoPROC will extract both cell and spacegroup from the information in the MTZ header

Consistent indexing for specific spacegroups

A good overview and introduction about possible alternative indexing schemes is given on the CCP4 documentation. To ensure that the autoPROC results are indexed consistently with a reference dataset, the –ref flag can be used:

% process -ref apo.mtz ...

Now not only the spacegroup and cell of that reference dataset will beused, but (if the spacegroup allows) the choice of indexing will be checked against the reference MTZ file. For that, the reference MTZ file need to contain a set of amplitude and sigma (often called F/SIGF or FP/SIGFP).

Usage of identical test-set flags for refinement.

If an existing (isomorphous) model is going to be used as the starting point of refinement against the newly obtained dataset, it is a good idea to carry the same test-set reflections over into the output files from autoPROC. The easiest way is to include a test-set flag column in the reference MTZ file to be used with the –ref command-line flag. This flag could have the column name FreeR_flag, FREE or FreeRflag:

% process -ref apo.mtz ...

will then take that test-set flag and add it to the output MTZ file (usually called truncate-unique.mtz or similar). If the current dataset diffracts to higher resolution than the reference MTZ file, the test-set flag will be complemented.