Dear Takaaki, I see, but I would say this is an text editing issue not a COOT failure. If I'm merging two chain from two separate pdb files together I don't want all the "crap" from two headers in one pdb file. So I would have to edit that new header anyway after merging - you just need to "copy-and paste" your previous LINK records from an pre-COOT file into the post-COOT file.
It is recommended jut to delete the entire pdb header above the CRYST card (maybe leaving link records in after reviewing them) , so that REFMAC is writing out a new one, as some information like gaps or links won't be updated in a new run. COOT does not know anything about links in a given PDB file. It even cannot read the LINKR record from REFMAC, so you have to delete the "R" from the REFMAC output and name it "LINK", but this will give you only the dotted line in COOT - this is not a restrain in COOT ! Cheers Stefan -----Ursprüngliche Nachricht----- Von: Mailing list for users of COOT Crystallographic Software [mailto:[email protected]] Im Auftrag von Takaaki Fukami Gesendet: Mittwoch, 18. Januar 2012 10:35 An: [email protected] Betreff: Re: N-terminal ACE restraints Dear Stefan, > You have to specify link records yourself to be used by REFMAC. Yes, I understand it. There were no LINK records in a pdb file written by Coot for refmac refinement, that was the reason of lacking restraints for ACE. In this particular case, there were 2 chains, A and B, and I stored them in separate files. Only chain B contained an ACE residue and a LINK record. I loaded the two files, and merged these chains in Coot before refmac refinement. I merged B into A, which seems to be a cause of the lacking LINKs in the refmac-input.pdb. I think these LINK information should be kept after merging chains... Thank you Takaaki -----Original Message----- From: Stefan Gerhardt [mailto:[email protected]] Sent: Wednesday, January 18, 2012 5:42 PM To: 深海隆明(探索研究部蛋白��造解析Ｇ); [email protected] Subject: AW: N-terminal ACE restraints Dear Takaai You have to specify link records yourself to be used by REFMAC. Use JLIGAND for link records see http://www.ysbl.york.ac.uk/mxstat/JLigand/tutorial_link.html Cheers Stefan -----Ursprüngliche Nachricht----- Von: Mailing list for users of COOT Crystallographic Software [mailto:[email protected]] Im Auftrag von Takaaki Fukami Gesendet: Mittwoch, 18. Januar 2012 08:00 An: [email protected] Betreff: Re: N-terminal ACE restraints Dear Paul, Coot handles the ACE group properly, though it won't write any LINK record to a pdb file to refine it by refmac. Therefore, the ACE group moved away in the refined structure. Could you make Coot to write out the LINK records at leaset present in an input pdb file? Regards, Takaaki -----Original Message----- From: Paul Emsley [mailto:[email protected]] Sent: Tuesday, January 17, 2012 9:47 PM Subject: Re: N-terminal ACE restraints On 17/01/12 12:01, Takaaki Fukami wrote: > Dear Coot experts, > > I have a protein with an N-acetyl group at its N-terminal, and want to > refine it with coot & refmac5. > Refmac recognize the LINK and applied restraints in $CLIB/list/mon_lib_list.cif. > > In Coot, when I "refine zone" from B0 to B1, the ACE group moved away > from the next PHE. They won't connect each other. Try sphere refine, that has more sophisticated linking rules. > > I extracted the data_link_ACE_C-N section from the mon_lib_list.cif > file and import it into Coot, but it didn't work. > > How do I include the link restraints into Coot? > It should Just Work. If it doesn't then I need to fix it. I've added this to my list. LINKRs for refinement are scheduled for 0.8. Regards, Paul.
