On 13/01/2017 03:29, Buth Sergii wrote:
However, they have to be modeled properly. The best option to model these sites
is to split
them into alternative conformations and place accordingly to chemical
environment. But as
soon as I split a given Zn residue and attempted to real space refine its parts
in linear
mode the Coot prompts “No restrains found” error message. Sphere refinement
does the job… by
putting alt confs together, basically overlaying them. For alt confs of often
occurring
residues like Na, Cl, Mg real space refinement works fine in the linear mode,
whereas in
sphere mode it acts as for Zn – overlays alt confs.
That is a result of the minimization technique - it's generally a useful short-cut ...
except in cases like these.
Is there a way to make Coot understand and refine alt confs of Zn or any other
rare
occurring residues? Making refinement operational in linear and sphere mode
would be a great
advantage.
Until I implement a more robust algorithm, there's a work around. Use the Fixed Atoms
(under the anchor icon) to fix an atom in the residue - it will be marked in green and not
move when refinement is activated and that will stop the superposition. You can move fixed
atoms by control-dragging on them.
And the last, but very important. Sphere refinement in Coot becomes very useful
when
treating my Zn sites.
Yes.
However, I wonder where spatial restrains come from?
Well, today Coot uses values from ener_lib.cif in vdW distance restraints. You can add
external distance restraints if you like.
We know that refinement of metals and their environments in Coot (and, if I may be so bold,
Refmac) is not all it could be and are actively working to improve this.
Any library like ReadySet uses?
I don't know what ReadySet does for metals.
> Restrains, generated by ReadySet cannot be used in Coot.
Depending on the nature of the restraints, this is something that the ReadySet developers
and I might be able to fix rather straightforwardly.
I run Coot 0.8.8-pre,
:-) good for you.
Paul.
