On Fri, 6 Mar 2015, Andy Howard wrote:
I'm refining a structure of a protein for which the crystallization
conditions include Na/K tartrate. The "Modeling -> Add Other Solvent
Molecules -> Add a new residue" menu in Coot recognized TAR as being
tartaric acid, so I used it to build this molecule. But this is a
high-resolution structure wherein I can discern the actual
stereochemistry of this molecule, and it's clearly not the one that the
code is building: the code is building (2R,3S)-dihydroxybutanedioic
acid, i.e. meso-tartaric acid, whereas the true solvent is the natural,
(R,R)-tartaric acid. The label within Coot says it's building (2S,3S),
but I don't think that's actually what's being coded for. I was able to
manually move the incorrect hydroxyl by "Rotate/Translate Zone", but of
course Coot then complains that there's an incorrect chiral volume. A
related peculiarity is that when I invoke the "Add other solvent
molecules" menu, it sometimes labels this compound as TARTARIC ACID, and
other times as "2,3-dihydroxybutanedioc acid". When it labels it as
TARTARIC ACID, it knows what chi angles are available; when it labels it
as "2,3-dihydroxybutanedioc acid", the "Edit Chi Angles" functionality
fails. I'm not an organic chemist, and I don't play one on television.
Any thoughts from the community?
Andy,
I do not know anything about adding solvent molecules in coot but thought
that it might be useful to get some information about the different
structural isomers of tartaric acid in the PDB chemical components
definitions (so that if/when you deposit the structure you will be
using the correct 3 letter code).
Searching Ligand expo http://ligand-expo.rcsb.org
for "Molecular name (similar)" yields the results that there are versions
of tartaric acid in the chemical components:
+=====+=======================================+
| PDB | Name/
| ID | Synonyms
+=====+=======================================+
| TAR | d(-)-tartaric acid
| | (2s,3s)-2,3-dihydroxybutanedioic acid
+=====+=======================================+
| TLA | l(+)-tartaric acid
| | (2r,3r)-2,3-dihydroxybutanedioic acid
+=====+=======================================+
| SRT | s,r meso-tartaric acid
| | (2r,3s)-2,3-dihydroxybutanedioic acid
| | (2s,3r)-2,3-dihydroxybutanedioic acid
+=====+=======================================+
Using http://www.rcsb.org/ additional information
TAR is found in 93 PDB entries
TLA is found in 275 PDB entries
SRT is found in 49 PDB entries
You can buy all three forms from Sigma Aldrich
So it sounds to me that you want to build your
tartaric acid(s) as TLA?
In case it helps I generated a dictionary for TLA on the Grade Web Server:
http://grade.globalphasing.org/cgi-bin/grade/server.cgi?runid=14257499237feLtSSYKZB
The cif dictionary should work with coot (and refmac). (But you might
prefer to use ccp4 dictionary and tools).
Good luck
Oliver
------------------
Dr Oliver Smart
Director SmartSci Limited
http://www.smartsci.uk/
& Consultant Global Phasing Ltd
http://www.globalphasing.com/
