Print

---

Hi Pavel

I work with peptide-scaffold chimeras on a daily basis. Initially, I tried to include the scaffolds as HETATM and use JLigand to create the links between the scaffolds and the natural amino acids. This procedure however is rather fussy and by now I build the chimera in silico, using natural amino acids provided by the program. Subsequently, I run the energy minimization on my own to have a rough control over the conformation. I then feed the coordinates into PRODRG without any further minimization and use the pdb and the restraints as they come from the server. 

Of course, such a ligand will not make use of the cif libraries of the natural amino acids during refinement. But if you imported the natural amino acids while building the chimera, your minimization used the proper restraints. I have the impression that this procedure is more efficient than trying to maintain the LINK in the pdb header (which, in my case, oftern got lost after running a refinement, so I had to copy-paste it anew after every cycle).

Best, Matthias

From: CCP4 bulletin board <[log in to unmask]> on behalf of Pavel Mader <[log in to unmask]>
Sent: Monday, November 4, 2019 6:06:11 PM
To: [log in to unmask]
Subject: Re: [ccp4bb] why does Coot ignore CONECT?

Hi Paul,

thank you for your answer. My problem is that I am building a cyclic peptide with non-standard amino acids, the side chains of which are linked by click chemistry... I know how to make modified amino acids that will make peptide bonds with their neighbors in the polypeptide chain, but I don't know how to make a covalent link between side chains of amino acids say No. 3 and 12 for example (of a 15 AA long chain). I was trying to solve the issue by making a CONECT record in the pdb file, but Coot ignored it. LINK "sort of" worked, but it was not possible to control the geometry (bond length and angles), often the real space refinement simply "explodes" probably in a attempt to avoid clashes. My current workaround is to make a cif file for the whole 15 AA long polypeptide (the bonds and angles now behave more or less as expected), but I don't consider this as a good general practice for building long polypeptide chains...

Thanks in advance for any hints guiding me in the right direction.

Pavel

[log in to unmask]">On 01/11/2019 20:17, Pavel Mader wrote:
> Hello,

Hello.

>
> I have a question, can anyone explain, why does Coot not display a covalent bond manually specified by
> CONECT line in a pdb file?

Because I have never thought them necessary. Using SSBOND, LINK and residue dictionaries seemed to me to
cover the bases for which CONECT would be used.

Paul.

########################################################################

To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1

---

To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB&A=1