Dear All,
I have received the following suggestion from a crystallographer. The
bottom part of the email was indicating that the mail was confidential,
so I have removed the name of the author of that email (together with
this section about confidentiality) and I am therefore forwarding the
message below.
What worked for me was to temporarily renumber residue 105 to 104 (this
I did with a text editor, not with Coot), fix the bond by real-space
refinement, and immediately change residue 104 back to 105. I did not
need to have a LINK card in the PDB file. That did the trick, so I
thought I should inform the bb even though it is "confidential".
Fred.
Dear Fred,
Here’s what you should do:
1. Briefly renumber your protein to sequential numbering (103, 104
etc.) in Coot.
2. Use real space refine zone in coot to tidy up the peptide bond
between residue 103 and 104.
3. When it looks good renumber the residues back to the standard
numbers (103, 105 etc.) in Coot
4. Add this line to your PDB header:
LINKR C TRP A 103 N ASN A 105 TRANS
5. Refine in Refmac. There will be no problem with it breaking the
bond any more because the TRANS word tells it to link the two
amino acids together as normal.
6. Bob’s your uncle!
Hope this helps.
|