Hello everybody,
I am working with a protein that should be covalently bound to an
organic ligand and I would calculate the correlation coefficient of this
ligand both against 2Fo-Fc and Fo-Fc map arising from refining (map
calculated in the presence of ligand).
To do this, I am using OVERLAPMAP (in CCP4, graphic interface) by
setting:
List a correlation by "residue"
First map "calculated from input MTZ"
F1 "FOFCWT" or "2FOFCWT"
PHI "PHFOFCWT" or "PH2FOFCWT"
sigma unassigned
weight unassigned
Second map "calculated from input coordinates" (this input
contains only the ligand molecule)
In addition, I am using PHENIX.GET_CC_MTZ_PDB (in phenix, command line)
to compare the previous results. In order to do the same kind of
calculation performed by OVERLAPMAP, I run:
phenix.get_cc_mtz_pdb *.pdb *.mtz \ labin="FP=FOFCWT PHIB=PHFOFCWT" \
atom_selection="chain a"
phenix.get_cc_mtz_pdb *.pdb *.mtz \ labin="FP=2FOFCWT PHIB=PH2FOFCWT" \
atom_selection="chain a"
("chain a" contains only the ligand)
My questions are:
1) OVERLAPMAP: Are the setting correct to perform CC calculations of a
model against Fo-Fc and 2Fo-Fc maps?
2) OVERLAPMAP: By considering that is not possible to define main-chain
and side-chain for the ligand molecule, is "Total correlation
coefficient", at the end of the output, the proper CC value of the whole
ligand?
3) PHENIX.GET_CC_MTZ_PDB: Is CC value, at the end of the output, the
equivalent of "Total correlation coefficient" arising from OVERLAPMAP?
3) Are the two procedures (ccp4 and phenix) similar or they perform
different calculations?
Thank for your suggestions.
