group on special position

Arthur Glasfeld

12 Aug 2020 12 Aug '20

5:53 p.m.

I’ve run into a situation where I think I have a phosphate sitting on a 3-fold axis, where the P to O1 bond lies on the axis and the atoms O2, O3 and O4 are in symmetry-related positions. The data go to 1.6 Å, so it looks pretty convincing. I’m refining this in Phenix and have tried some tricks with occupancy - fixing the atoms at 0.33, but I get a lot of positive density in the Fo-Fc map and large B-factors. I’ve looked around but am not finding ideas on how to model this correctly in my structure file - the references I find are for monoatomic ions. Any advice would be welcome. Thanks, Arthur ***************** Arthur Glasfeld Reed College Portland, OR 97202 Office: (503)517-7679

Show replies by date

Dr. Karine Röwer

12 Aug 12 Aug

10:37 p.m.

Hi Arthur, from the way I used to work in small molecule crystallography, I'd set the occupancy of the atoms on the symmetry axis to 0.33, and keep only one of the other three O atoms with an occupancy of 1. I'd expect phenix to generate the other two O atoms via symmetry operators, and applied to the atoms on the symmetry axis, you should get a total occupancy of 1. Does that help? Cheers, Karine On Thu, Aug 13, 2020 at 2:56 AM Arthur Glasfeld wrote:

...

I’ve run into a situation where I think I have a phosphate sitting on a 3-fold axis, where the P to O1 bond lies on the axis and the atoms O2, O3 and O4 are in symmetry-related positions. The data go to 1.6 Å, so it looks pretty convincing. I’m refining this in Phenix and have tried some tricks with occupancy - fixing the atoms at 0.33, but I get a lot of positive density in the Fo-Fc map and large B-factors. I’ve looked around but am not finding ideas on how to model this correctly in my structure file - the references I find are for monoatomic ions. Any advice would be welcome.

Thanks, Arthur

***************** Arthur Glasfeld Reed College Portland, OR 97202 Office: (503)517-7679

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

Arthur Glasfeld

13 Aug 13 Aug

10:18 a.m.

Hi Karine, I did give that a try, but the phosphate jumped out of the density. My guess is that the symmetry relationship between O2, O3 and O4 means that if each is at full occupancy, they then occupy each of the three positions so there’s a clash between all three in each asymmetric unit. I’ve also tried setting O2 to 1.00 occupancy, and O3 & O4 to 0.00. Again, the phosphate jumps out of density... Thanks, Arthur

...

On Aug 12, 2020, at 10:37 PM, Dr. Karine Röwer wrote:

Hi Arthur, from the way I used to work in small molecule crystallography, I'd set the occupancy of the atoms on the symmetry axis to 0.33, and keep only one of the other three O atoms with an occupancy of 1. I'd expect phenix to generate the other two O atoms via symmetry operators, and applied to the atoms on the symmetry axis, you should get a total occupancy of 1. Does that help? Cheers, Karine

On Thu, Aug 13, 2020 at 2:56 AM Arthur Glasfeld wrote: I’ve run into a situation where I think I have a phosphate sitting on a 3-fold axis, where the P to O1 bond lies on the axis and the atoms O2, O3 and O4 are in symmetry-related positions. The data go to 1.6 Å, so it looks pretty convincing. I’m refining this in Phenix and have tried some tricks with occupancy - fixing the atoms at 0.33, but I get a lot of positive density in the Fo-Fc map and large B-factors. I’ve looked around but am not finding ideas on how to model this correctly in my structure file - the references I find are for monoatomic ions. Any advice would be welcome.

Thanks, Arthur

***************** Arthur Glasfeld Reed College Portland, OR 97202 Office: (503)517-7679

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

Pavel Afonine

11:12 a.m.

Hi Arthur, hi Karine, this is definitely possible to do in Phenix but off the top of my head I can't remember how. What you have tried so far sounds about right to me but there might be subtle nuances. Would you mind sending me the model file so that I figure out the details and reply back with a how-to recipe? I'll use this as an opportunity to improve the documentation too! I don't need the data file but if you share it too then I'll make sure the refinement works fine with whatever setup we'll come up with. All files will be treated confidentially. Please send files to me only, not the whole phenixbb! Pavel On 8/13/20 10:18, Arthur Glasfeld wrote:

...

Hi Karine, I did give that a try, but the phosphate jumped out of the density. My guess is that the symmetry relationship between O2, O3 and O4 means that if each is at full occupancy, they then occupy each of the three positions so there’s a clash between all three in each asymmetric unit. I’ve also tried setting O2 to 1.00 occupancy, and O3 & O4 to 0.00. Again, the phosphate jumps out of density...

Thanks, Arthur

...
On Aug 12, 2020, at 10:37 PM, Dr. Karine Röwer wrote:

Hi Arthur, from the way I used to work in small molecule crystallography, I'd set the occupancy of the atoms on the symmetry axis to 0.33, and keep only one of the other three O atoms with an occupancy of 1. I'd expect phenix to generate the other two O atoms via symmetry operators, and applied to the atoms on the symmetry axis, you should get a total occupancy of 1. Does that help? Cheers, Karine

On Thu, Aug 13, 2020 at 2:56 AM Arthur Glasfeld wrote: I’ve run into a situation where I think I have a phosphate sitting on a 3-fold axis, where the P to O1 bond lies on the axis and the atoms O2, O3 and O4 are in symmetry-related positions. The data go to 1.6 Å, so it looks pretty convincing. I’m refining this in Phenix and have tried some tricks with occupancy - fixing the atoms at 0.33, but I get a lot of positive density in the Fo-Fc map and large B-factors. I’ve looked around but am not finding ideas on how to model this correctly in my structure file - the references I find are for monoatomic ions. Any advice would be welcome.

Thanks, Arthur

Pavel Afonine

17 Aug 17 Aug

7:22 p.m.

For the archive record, to make this work: - set occupancy of PO4 on special position to 0.33; - use custom_nonbonded_symmetry_exclusions="chain B and resseq 5" (which selects that PO4) to make sure it does not clash with its symmetry copies. Pavel On 8/13/20 10:18, Arthur Glasfeld wrote:

...

Hi Karine, I did give that a try, but the phosphate jumped out of the density. My guess is that the symmetry relationship between O2, O3 and O4 means that if each is at full occupancy, they then occupy each of the three positions so there’s a clash between all three in each asymmetric unit. I’ve also tried setting O2 to 1.00 occupancy, and O3 & O4 to 0.00. Again, the phosphate jumps out of density...

Thanks, Arthur

...
On Aug 12, 2020, at 10:37 PM, Dr. Karine Röwer wrote:

Hi Arthur, from the way I used to work in small molecule crystallography, I'd set the occupancy of the atoms on the symmetry axis to 0.33, and keep only one of the other three O atoms with an occupancy of 1. I'd expect phenix to generate the other two O atoms via symmetry operators, and applied to the atoms on the symmetry axis, you should get a total occupancy of 1. Does that help? Cheers, Karine

On Thu, Aug 13, 2020 at 2:56 AM Arthur Glasfeld wrote: I’ve run into a situation where I think I have a phosphate sitting on a 3-fold axis, where the P to O1 bond lies on the axis and the atoms O2, O3 and O4 are in symmetry-related positions. The data go to 1.6 Å, so it looks pretty convincing. I’m refining this in Phenix and have tried some tricks with occupancy - fixing the atoms at 0.33, but I get a lot of positive density in the Fo-Fc map and large B-factors. I’ve looked around but am not finding ideas on how to model this correctly in my structure file - the references I find are for monoatomic ions. Any advice would be welcome.

Thanks, Arthur

***************** Arthur Glasfeld Reed College Portland, OR 97202 Office: (503)517-7679

_______________________________________________ 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]

Arthur Glasfeld

4 Sep 4 Sep

2:22 p.m.

Hi Pavel, Sorry for the delay in testing this out. It worked fine! I added one additional measure, fixing occupancy at 0.33 so it wouldn’t refine with the other partial occupancies in my model. Thanks so much for your help, Arthur

...

On Aug 17, 2020, at 7:22 PM, Pavel Afonine wrote:

For the archive record, to make this work:

- set occupancy of PO4 on special position to 0.33; - use

custom_nonbonded_symmetry_exclusions="chain B and resseq 5"

(which selects that PO4) to make sure it does not clash with its symmetry copies.

Pavel

On 8/13/20 10:18, Arthur Glasfeld wrote:

...
Hi Karine, I did give that a try, but the phosphate jumped out of the density. My guess is that the symmetry relationship between O2, O3 and O4 means that if each is at full occupancy, they then occupy each of the three positions so there’s a clash between all three in each asymmetric unit. I’ve also tried setting O2 to 1.00 occupancy, and O3 & O4 to 0.00. Again, the phosphate jumps out of density...

Thanks, Arthur

...
On Aug 12, 2020, at 10:37 PM, Dr. Karine Röwer wrote:

Hi Arthur, from the way I used to work in small molecule crystallography, I'd set the occupancy of the atoms on the symmetry axis to 0.33, and keep only one of the other three O atoms with an occupancy of 1. I'd expect phenix to generate the other two O atoms via symmetry operators, and applied to the atoms on the symmetry axis, you should get a total occupancy of 1. Does that help? Cheers, Karine

On Thu, Aug 13, 2020 at 2:56 AM Arthur Glasfeld wrote: I’ve run into a situation where I think I have a phosphate sitting on a 3-fold axis, where the P to O1 bond lies on the axis and the atoms O2, O3 and O4 are in symmetry-related positions. The data go to 1.6 Å, so it looks pretty convincing. I’m refining this in Phenix and have tried some tricks with occupancy - fixing the atoms at 0.33, but I get a lot of positive density in the Fo-Fc map and large B-factors. I’ve looked around but am not finding ideas on how to model this correctly in my structure file - the references I find are for monoatomic ions. Any advice would be welcome.

Thanks, Arthur

***************** Arthur Glasfeld Reed College Portland, OR 97202 Office: (503)517-7679

_______________________________________________ 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]

Pavel Afonine

2:25 p.m.

Hi Arthur, thanks for confirming! What if that ion on special position is partially occupied? In that case you may want to refine its occupancy.. But clearly it's going to be tricky to decouple partial occupancy due to symmetry and the one due to disorder. Pavel On 9/4/20 14:22, Arthur Glasfeld wrote:

...

Hi Pavel, Sorry for the delay in testing this out. It worked fine! I added one additional measure, fixing occupancy at 0.33 so it wouldn’t refine with the other partial occupancies in my model.

Thanks so much for your help,

Arthur

...
On Aug 17, 2020, at 7:22 PM, Pavel Afonine wrote:

For the archive record, to make this work:

- set occupancy of PO4 on special position to 0.33; - use

custom_nonbonded_symmetry_exclusions="chain B and resseq 5"

(which selects that PO4) to make sure it does not clash with its symmetry copies.

Pavel

On 8/13/20 10:18, Arthur Glasfeld wrote:

...
Hi Karine, I did give that a try, but the phosphate jumped out of the density. My guess is that the symmetry relationship between O2, O3 and O4 means that if each is at full occupancy, they then occupy each of the three positions so there’s a clash between all three in each asymmetric unit. I’ve also tried setting O2 to 1.00 occupancy, and O3 & O4 to 0.00. Again, the phosphate jumps out of density...

Thanks, Arthur

...
On Aug 12, 2020, at 10:37 PM, Dr. Karine Röwer wrote:

Hi Arthur, from the way I used to work in small molecule crystallography, I'd set the occupancy of the atoms on the symmetry axis to 0.33, and keep only one of the other three O atoms with an occupancy of 1. I'd expect phenix to generate the other two O atoms via symmetry operators, and applied to the atoms on the symmetry axis, you should get a total occupancy of 1. Does that help? Cheers, Karine

On Thu, Aug 13, 2020 at 2:56 AM Arthur Glasfeld wrote: I’ve run into a situation where I think I have a phosphate sitting on a 3-fold axis, where the P to O1 bond lies on the axis and the atoms O2, O3 and O4 are in symmetry-related positions. The data go to 1.6 Å, so it looks pretty convincing. I’m refining this in Phenix and have tried some tricks with occupancy - fixing the atoms at 0.33, but I get a lot of positive density in the Fo-Fc map and large B-factors. I’ve looked around but am not finding ideas on how to model this correctly in my structure file - the references I find are for monoatomic ions. Any advice would be welcome.

Thanks, Arthur

***************** Arthur Glasfeld Reed College Portland, OR 97202 Office: (503)517-7679

_______________________________________________ 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]

1793

Age (days ago)

1815

Last active (days ago)

List overview

Download

6 comments

3 participants

participants (3)

Arthur Glasfeld
Dr. Karine Röwer
Pavel Afonine