on ligand restraint - phenixbb - phenix-online.org

newer
Special position ambiguity

on ligand restraint

older
Selection of Torsion NCS restraint...

Smith Lee

4 Oct 2015 4 Oct '15

9:19 a.m.

Dear All, Does phenix.refine can process some ligands without the process by eLBOW, for example ATP, ADP,etc? Smith

Reply

Sign in to reply online Use email software

Show replies by date

Pavel Afonine

4 Oct 4 Oct

6:14 p.m.

Yes. Pavel On 10/4/15 02:19, Smith Lee wrote:

Dear All,

Does phenix.refine can process some ligands without the process by eLBOW, for example ATP, ADP,etc?

Smith _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

Reply

Sign in to reply online Use email software

Smith Liu

5 Oct 5 Oct

4:51 a.m.

Dear All, I have tried a phenix.real_space_refine with a ligand cif and without ligand cif (both with ligand PDB in the whole protein-ligand PDB for refine), I find with ligand cif in the input can lead to significant increase of the clash score for the refined PDB. Will you please make comment on whether the clash score incensement by adding cif in the input was a general phenomenon, or only specific to my situation, and how to solve it? Best regards. Smith At 2015-10-05 02:14:28, "Pavel Afonine" wrote:

Yes. Pavel

On 10/4/15 02:19, Smith Lee wrote:

...
Dear All,

Does phenix.refine can process some ligands without the process by eLBOW, for example ATP, ADP,etc?

Smith _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

Reply

Sign in to reply online Use email software

Pavel Afonine

6:12 a.m.

Most likely wrong ligand CIF file. Pavel On 10/4/15 21:51, Smith Liu wrote:

Dear All,

I have tried a phenix.real_space_refine with a ligand cif and without ligand cif (both with ligand PDB in the whole protein-ligand PDB for refine), I find with ligand cif in the input can lead to significant increase of the clash score for the refined PDB.

Will you please make comment on whether the clash score incensement by adding cif in the input was a general phenomenon, or only specific to my situation, and how to solve it?

Best regards.

Smith

At 2015-10-05 02:14:28, "Pavel Afonine" wrote:

...
Yes. Pavel

On 10/4/15 02:19, Smith Lee wrote:

...
Dear All,

Does phenix.refine can process some ligands without the process by eLBOW, for example ATP, ADP,etc?

Smith _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

Reply

Sign in to reply online Use email software

Smith Liu

10:31 a.m.

Dear Pavel, My cif file was for ATP, which was rather easily got by eLBOW as for the Phenix contains it in its library. Attached is the ATP.cif file I used. In addition, I have run another 10 cycles of simulated_annealing+minimization_global, and the clash score was still 120 (too high). Recently I fond for both the 1.9-1692 and 1.10.2155 version of phenix.real_spacea_refine, I got rather high clash score (higher than 20). In addition I found (recently?) the phenix.real_space_refine output PDB clash score was different from the Ramachandran clash score value. Is an possibility my Windows versions of phenix have something wrong? I am looking forward to getting your further comment. Smith At 2015-10-05 14:12:39, "Pavel Afonine" wrote: Most likely wrong ligand CIF file. Pavel On 10/4/15 21:51, Smith Liu wrote: Dear All, I have tried a phenix.real_space_refine with a ligand cif and without ligand cif (both with ligand PDB in the whole protein-ligand PDB for refine), I find with ligand cif in the input can lead to significant increase of the clash score for the refined PDB. Will you please make comment on whether the clash score incensement by adding cif in the input was a general phenomenon, or only specific to my situation, and how to solve it? Best regards. Smith At 2015-10-05 02:14:28, "Pavel Afonine" wrote:

Yes. Pavel

On 10/4/15 02:19, Smith Lee wrote:

...
Dear All,

Does phenix.refine can process some ligands without the process by eLBOW, for example ATP, ADP,etc?

Smith _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

Reply

Sign in to reply online Use email software

Nigel Moriarty

4:45 p.m.

Smith

From the comments in the file, you generated the ATP restraints with 1.9-1692. I redid the command with 1.10-2155 and the restraints were better. Clearly, the program has improved in the 13 months that 1.9 was released and you should stop using it in preference for 1.10.

Cheers Nigel --- Nigel W. Moriarty Building 33R0349, Physical Biosciences Division Lawrence Berkeley National Laboratory Berkeley, CA 94720-8235 Phone : 510-486-5709 Email : [email protected] Fax : 510-486-5909 Web : CCI.LBL.gov On Mon, Oct 5, 2015 at 3:31 AM, Smith Liu wrote:

Dear Pavel,

My cif file was for ATP, which was rather easily got by eLBOW as for the Phenix contains it in its library. Attached is the ATP.cif file I used.

In addition, I have run another 10 cycles of simulated_annealing+minimization_global, and the clash score was still 120 (too high).

Recently I fond for both the 1.9-1692 and 1.10.2155 version of phenix.real_spacea_refine, I got rather high clash score (higher than 20). In addition I found (recently?) the phenix.real_space_refine output PDB clash score was different from the Ramachandran clash score value. Is an possibility my Windows versions of phenix have something wrong?

I am looking forward to getting your further comment.

Smith

At 2015-10-05 14:12:39, "Pavel Afonine" wrote:

Most likely wrong ligand CIF file. Pavel

On 10/4/15 21:51, Smith Liu wrote:

Dear All,

I have tried a phenix.real_space_refine with a ligand cif and without ligand cif (both with ligand PDB in the whole protein-ligand PDB for refine), I find with ligand cif in the input can lead to significant increase of the clash score for the refined PDB.

Will you please make comment on whether the clash score incensement by adding cif in the input was a general phenomenon, or only specific to my situation, and how to solve it?

Best regards.

Smith

At 2015-10-05 02:14:28, "Pavel Afonine" wrote:

...
Yes. Pavel

On 10/4/15 02:19, Smith Lee wrote:

...
Dear All,

Does phenix.refine can process some ligands without the process by eLBOW, for example ATP, ADP,etc?

Smith _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

Reply

Sign in to reply online Use email software

Smith Lee

4 p.m.

New subject: Fw: on ligand restraint

Dear Pavel, I find some reasons for the high clash. If the protein part PDB is without H and the ATP PDB part has H, the final clash score will be high. But it seems cif does not permit ligand without H, am I right? If I add H to both the protein and ligand, the poor resolution (poorer than 4) may be does not like to have the H added. What is your opinion? Will you please explain whether the phenix.real_space_refined pdb's clashing score is same with that of MolProbity? If same, then there may be something wrong with my phenix software, as for recently I have 2 proteins, with phenix.real_space_refine, the critical clashing score 20 level cannot be got (always higher than 20). Best regards. Smith On Monday, October 5, 2015 2:52 PM, Smith Liu wrote: Dear All, I have tried a phenix.real_space_refine with a ligand cif and without ligand cif (both with ligand PDB in the whole protein-ligand PDB for refine), I find with ligand cif in the input can lead to significant increase of the clash score for the refined PDB. Will you please make comment on whether the clash score incensement by adding cif in the input was a general phenomenon, or only specific to my situation, and how to solve it? Best regards. Smith At 2015-10-05 02:14:28, "Pavel Afonine" wrote:

Yes. Pavel

On 10/4/15 02:19, Smith Lee wrote:

...
Dear All,

Does phenix.refine can process some ligands without the process by eLBOW, for example ATP, ADP,etc?

Smith _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb Unsubscribe: [email protected]

Reply

Sign in to reply online Use email software

3371

Age (days ago)

3372

Last active (days ago)

Download

6 comments

4 participants

tags

participants (4)

Nigel Moriarty
Pavel Afonine
Smith Lee
Smith Liu