Eleanor Dodson wrote:
Offering a selection of space groups compatible with a point group will
not require reindexing.
Some point groups of course permit reindexing, but one hope you have
chosen a consistent set as you combine data sets into an MTZ file -
pointless and the GUI both offer ways to check this.
Phasing cannot be carried out till you have chosen a spacegroup, and any
existing phase associated data should probably be weeded out as part of
a "Assign Spacegroup" function.
Good, that makes sense.
Note however we have introduced a constraint on the data model:
- Phasing and multiple spacegroups are inconsistent.
Actually, IIUC, it's more complicated, since heavy atom location comes
before phasing, but produces phases with enantiomorph ambiguity.
Grahame's 4-step indicator of uncertainty provides the right sort of
framework. But each step is associated with a corresponding set of
restrictions on what is valid in the file - and in the case of phases,
what they actually mean.
