Just for info:
There is new version of the dictionary that is compliant with pdb v3.
It works with new version of refmac and should work with phenix also.
It seemed to work with coot also.
www.ysbl.york.ac.uk/refmac/data/refmac_experimental/remfac_dictionary_v5.18.tar.gz
I should warn that we are still working on this dictionary but it
should be usable now (if you use refmac then you will need newer
version of refmac also available from the same page)
regards
Garib
On 16 Mar 2010, at 19:56, Paul Emsley wrote:
Martin P. Horvath wrote:
I am refining with 1.14 Å data and have included riding hydrogen
atoms. In
addition to (and perhaps related to) certain difficulties with real-
space
refinement caused by nomenclature mismatch phenix/coot, the alpha-
carbon
hydrogen is not removed when deleting an alternate conformation
residue. For
each residue previously with two alternate conformations, reduced
to one
conformation, the H atom retains A and B tags with partial
occupancy. This
is easy to clean up with a text editor, or by running phenix.reduce
-trim;
phenix.ready_set, however, if there are plans to smooth out model
building
with hydrogens...
OK, I should fix that right away.
As an aside, re: the dictionary in general:
Coot is distributed with a copy of the refmac library, which has PDB
version 2.3 atom names.
You can override that dictionary by setting up CCP4 in the normal way.
You can further override the dictionary by setting the environment
variable COOT_REFMAC_LIB_DIR
So if Phenix distributes a version of the monomer library that has
PDB version 3.x atom names then Coot can use that if you point to it.
Paul.
