Hello, I am looking for advice on how to best refine the structure of a complex that has one component that can adopt two completely different conformations (a DNA duplex). I can model both conformations and set occupancies to 0.5, and try to constrain_group their occupancies, but in phenix.refine the two conformations are interpreted as major clashes. Thanks. Best, John -- John Pascal, PhD ph 215.503.4596 Assistant Professor fx 215.923.2117 Thomas Jefferson University Biochemistry & Molecular Biology 233 South 10th Street, BLSB 804 Philadelphia, Pennsylvania 19107 ---------------------------------------------------------------- This message was sent using IMP, the Internet Messaging Program.
Hi John,
I am looking for advice on how to best refine the structure of a complex that has one component that can adopt two completely different conformations (a DNA duplex). I can model both conformations and set occupancies to 0.5, and try to constrain_group their occupancies, but in phenix.refine the two conformations are interpreted as major clashes.
The trick to avoid the clashes is to set the "altloc" characters in column 17 of your PDB file. E.g. assign "A" to all atoms of the first conformer, and "B" to all atoms of the second. phenix.refine should correctly interpret your PDB file if you have a whole chain with altloc A followed by a whole chain with altloc B, assuming the chain id is identical and there are no TER cards between the conformers. Ralf
Hi Ralf, is this concept expandable to refining two or more ligands with different resnames and atomnames on top of each other in the same physical location using altlocs? I vaguely recall that phenix supports this approach but have not tested it out myself yet. Hypothetical case in point: 3 Ligands with resnames LG1, LG2 and LG3, differing atomnames, all chain C and resi 1, altlocs from A to C. Is something like this supported? What kind of requirements would I need in order to do this? Cheers, Carsten
-----Original Message----- From: [email protected] [mailto:phenixbb- [email protected]] On Behalf Of Ralf W. Grosse-Kunstleve Sent: Wednesday, December 02, 2009 3:55 PM To: [email protected] Subject: Re: [phenixbb] (no subject)
Hi John,
I am looking for advice on how to best refine the structure of a complex that has one component that can adopt two completely different conformations (a DNA duplex). I can model both conformations and set occupancies to 0.5, and try to constrain_group their occupancies, but in phenix.refine the two conformations are interpreted as major clashes.
The trick to avoid the clashes is to set the "altloc" characters in column 17 of your PDB file. E.g. assign "A" to all atoms of the first conformer, and "B" to all atoms of the second.
phenix.refine should correctly interpret your PDB file if you have a whole chain with altloc A followed by a whole chain with altloc B, assuming the chain id is identical and there are no TER cards between the conformers.
Ralf _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
Hi Carsten,
is this concept expandable to refining two or more ligands with different resnames and atomnames on top of each other in the same physical location using altlocs?
yes. An example in PDB is: http://www.rcsb.org/pdb/files/1EJG.pdb where residues #22 will be treated ok in phenix.refine automatically.
I vaguely recall that phenix supports this approach but have not tested it out myself yet.
I implemented this more than a year ago right after Ralf made it possible to get occupancy groupings as part of PDB interpretation.
Hypothetical case in point: 3 Ligands with resnames LG1, LG2 and LG3, differing atomnames, all chain C and resi 1, altlocs from A to C. Is something like this supported? What kind of requirements would I need in order to do this?
I'm not sure I understood this after spending 2 seconds looking at it, but if you send me an example PDB file and tell what is supposed to be coupled with what, then I will send you an example. (I'm leaving for Zürich right now and hopefully will be able to reply tonight). Pavel.
Pavel, 1EJG is an interesting case and pretty much reflects what I am thinking of. Instead of known aminoacids I am more thinking in the line of ligands but same idea. Glad to see that phenix can handle this case. Cheers, Carsten
-----Original Message----- From: [email protected] [mailto:phenixbb- [email protected]] On Behalf Of Pavel Afonine Sent: Thursday, December 03, 2009 8:57 AM To: PHENIX user mailing list Subject: Re: [phenixbb] (no subject)
Hi Carsten,
is this concept expandable to refining two or more ligands with different resnames and atomnames on top of each other in the same physical location using altlocs?
yes. An example in PDB is: http://www.rcsb.org/pdb/files/1EJG.pdb where residues #22 will be treated ok in phenix.refine automatically.
I vaguely recall that phenix supports this approach but have not tested it out myself yet.
I implemented this more than a year ago right after Ralf made it possible to get occupancy groupings as part of PDB interpretation.
Hypothetical case in point: 3 Ligands with resnames LG1, LG2 and LG3, differing atomnames, all chain C and resi 1, altlocs from A to C. Is something like this supported? What kind of requirements would I need in order to do this?
I'm not sure I understood this after spending 2 seconds looking at it, but if you send me an example PDB file and tell what is supposed to be coupled with what, then I will send you an example. (I'm leaving for Zürich right now and hopefully will be able to reply tonight).
Pavel.
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
Hi Carsten, ok, great ! Yes, ligands will be treated the same way as everything else in this case. Please let me know if you have any questions or problems with this and I will be happy to help. All the best! Pavel. On 12/3/09 11:27 AM, Schubert, Carsten [PRDUS] wrote:
Pavel,
1EJG is an interesting case and pretty much reflects what I am thinking of. Instead of known aminoacids I am more thinking in the line of ligands but same idea. Glad to see that phenix can handle this case.
Cheers,
Carsten
-----Original Message----- From: [email protected] [mailto:phenixbb- [email protected]] On Behalf Of Pavel Afonine Sent: Thursday, December 03, 2009 8:57 AM To: PHENIX user mailing list Subject: Re: [phenixbb] (no subject)
Hi Carsten,
is this concept expandable to refining two or more ligands with different resnames and atomnames on top of each other in the same physical location using altlocs?
yes. An example in PDB is: http://www.rcsb.org/pdb/files/1EJG.pdb where residues #22 will be treated ok in phenix.refine automatically.
I vaguely recall that phenix supports this approach but have not tested it out myself yet.
I implemented this more than a year ago right after Ralf made it possible to get occupancy groupings as part of PDB interpretation.
Hypothetical case in point: 3 Ligands with resnames LG1, LG2 and LG3, differing atomnames, all chain C and resi 1, altlocs from A to C. Is something like this supported? What kind of requirements would I need
in
order to do this?
I'm not sure I understood this after spending 2 seconds looking at it, but if you send me an example PDB file and tell what is supposed to be coupled with what, then I will send you an example. (I'm leaving for Zürich right now and hopefully will be able to reply tonight).
Pavel.
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
participants (4)
-
John Pascal -
Pavel Afonine -
Ralf W. Grosse-Kunstleve -
Schubert, Carsten [PRDUS]