Prodrg does very good job in guessing atom types and bond orders. Once
it guessed things right it does extremely good job in generating bond
lengths, angles etc.
But it still depends on accuracy of coordinates. You need to check
carefully bonding orders before using them in refinement.
You can also try JLigand available from:
www.ysbl.york.ac.uk/mxstat/
You need to click JLigand. There are tutorials how to use JLigand and
use them in coot also. One tutorial is for ligands and another
tutorial is for covalent links.
Regards
Garib
P.S. refmac, phenix.refine and coot use the same dictionary (buster/
tnt may have tools to convert or use them directly also)
On 21 Jun 2010, at 11:40, Paul Lindblom wrote:
> Hi all,
>
> I need to create a cif file for a new ligand that does not exist in
> the pdb, so far. Normally refmac created such a cif file when I
> merged the ligand to the structure, but this time it stops with a
> "fatal error". I think the alternatives are "sketcher" and the
> prodrg server. Can anybody tell me if I have to define the geometric
> restraints by myself, or is it possible to take just the output
> files and put them into refmac.? And if I have to define the
> geometric restraints, how to do this?
>
> Thanks,
>
> Paul
|