On Sat, Mar 3, 2012 at 8:27 AM, Pavel Afonine
phenix.real_space_correlation
allows to compute correlations between nearly any user-defined maps, can use x-ray or neutron scattering dictionaries, accepts atoms selections, detail level (cc per residue, atom etc).
This doesn't let you define your own level of detail, e.g. CC for a domain, or a loop. It would need to be refactored to pick out map grid points corresponding to a user-input atom selection instead of each residue or each atom. This could theoretically be done for get_cc_mtz_pdb, but I think that uses RESOLVE internally to handle the origin shift, which makes it non-trivial to change. In this case it is probably overkill.
I don't remember if it is available in the GUI (Nat to confirm), if not then it would be good to have.
It's not in the GUI because I found it far too confusing - looking at the code again I am still not sure what some of the parameters mean. I assume this is because it's written to be flexible enough to handle two models or two maps in addition to model-vs-map, but this makes it far too cumbersome for routine use. The validation GUI includes the same residue-or-atom enumeration anyway, which is what 99% of users would use phenix.real_space_correlation for anyway. What Francis is asking for is something like this: phenix.some_awesome_new_command model.pdb data.mtz selection="rename LIG" which phenix.real_space_correlation cannot do, especially since it does not seem to be able to figure out what to do with model and data files automatically. -Nat