Hi Yahui!

I am having this problem as well again and agin. Most problematic is it, if you have non-standard atoms in your compound.
I don't really know whrere the problem lies, but here is what I do:
Do not use sketcher!
Simply generate your ligand using coot by placing atoms into the density (you may start from standard compounds if available). Save your ligand in a pdb file and make sure (text editor), that all atoms belong to the same compound indicated by the same 3 letter identfier (e.g. LIG).
Merge this PDB file with your protein in Coot and save.
Run Refmac.
Refmac will abort but before stopping it will put out a library file with recommended bonds and angles and so on. This file you should manually edit putting in your chemical knowlege of the ligand.

Use this cif file in a second run for refmac and for all coot real space refinements.

Have fun,

Matthias

PS: sometimes (no non-standard atoms) a simpler way is to leave the field "Regularize with Refmac" unchecked, when you create your final library file.

Am 6/9/2010 12:23 PM, schrieb Yahui Yan:

Hello,

Could you please help me with the sketcher?

I'm trying to use ccp4 sketcher to generate a new ligand and then complex it with a protein in coot. I've drawn the ligand, numbered each atom and defined each bond type. Then I ran save file, create library description. The pdb file was loaded to coot and worked fine. Then I imported cif file. However, when I tried to refine the ligand, a message popped out, saying 'No restrains'. I double checked the numbering, I think it's Ok. Did I do anything wrong? I'm really struggling on this. If you need more information, please let me know. Thanks very much.

Best regards,
Yahui


--
****************************************************
Dr. Matthias Zebisch
Universität Leipzig
Biotechnologisch-Biomedizinisches Zentrum
Strukturanalytik von Biopolymeren
Deutscher Platz 5
04103 Leipzig
Germany
Phone: 0049-341-97-31323 (lab) -31312 (office)
Fax  : 0049-341-97-31319
email: [email protected]
****************************************************

Reply via email to