*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
I am confused now.
Here are my thoughts:
There is no ambiguity upon calculating Fcalc's. The symmetry is present
in the PDB file and is specified by the
unit cell parameters, and the standard setting of the space group. I
guess that if you compute and FFT of a molecule N times, N times you
will get the same set of SF's. if not, we might have a problem.
The ambiguity comes in to place only because the there are multiple
equivalent autoindexing solutions, whose resulting datasets (without
reindexing) have not equal initensities for their miller indices.
Simply using a moleculer model refined againt data set A to see if there
is a ligand in dataset B, is dangerous.
If you just take the model from your previous experiment, you are
basically indirectly comparing two datasets that you have separately
indexed. In this case you are of course having a 50% change of having it
wrong.
Solve this problem by molecular replacement, or use programs like
iotbx.reflection_statistics, pointless/reindex+sftools to sort this out.
In serous need of some cofee,
Peter
Ronald E Stenkamp wrote:
*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
If the only thing you're giving CNS is the pdb file, and you let it
generate the indices for the Fcs, there is likely no way to control
which orientation of the reciprocal lattice it'll come up with.
If in addition you give it a file with hkl values, the resulting
Fcs should be consistent with the axes defined by that file. There
must be something different about the way you're calculating Fcs if
that's not what you observe.
Ron
On Wed, 29 Mar 2006, William Scott wrote:
My main point is that if you simply take a pdb file from say a
hexagonal crystal, put it into a trigonal cell, use (eg) CNS to
calculate Fcalcs, there is no way to know in advance which way to
index a native data set that never sees or knows about any of this.
