Hi Francis, Thanks for your response. The matthews coefficient suggests two molecules in the AU. Phaser also finds two molecules. I ran the dataset through phenix xtriage it did not indicate twinning though. The molecule also exists in nature as a monomer. Shya

Twinning? What's your matthews coefficient say? Do you know if your structure is a multimer (biochemistry, etc)? Does it agree with the matthews coefficient?

If the unit cell is not big enough to hold all of the contents,then this is an indicator for twinning .

FR

On Jul 23, 2009, at 3:09 PM, [email protected] wrote:

...

Hi all,

I was trying to solve a structure by molecular replacement. I scaled the data in P4222 space group (resolution 2.7A) with two molecules in the assymmetric unit (molecule A and B) I ran phaser with my model and got a Zscore of 5.1. When I look at the map that I got from phaser I could easily see good electron density for both molecules, However upon inspection of the electron density map there were considerable interaction or clashes with molecule B and a symmetry atom. Molecule A had no clashes however with the symmetry atoms. I was wondering if anyone knows how to resolve this. Could it be a problem of space group. The statistics are good for space group P4222 and the I/sigI was good till 2.7A. Any advice is appreciated, Shya

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

--------------------------------------------- Francis Reyes M.Sc. 215 UCB University of Colorado at Boulder

gpg --keyserver pgp.mit.edu --recv-keys 67BA8D5D

8AE2 F2F4 90F7 9640 28BC 686F 78FD 6669 67BA 8D5D

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

Shya, Did phaser complain that the asymmetric unit was too full? How do the self rotation maps look? Are the crystallographic peaks exact or off by a few degrees (your resolution data may make it difficult to see this)? How do the N(z) cumulative intensity distributions look (make sure to calculate this with thin resolution bins, i.e. increase BINS in Scala I think)? Does your data look sigmoidal on this plot? Perfect twinning or an NCS that's close to a crystallographic axis is difficult to diagnose from merged intensity statistics and even more difficult with resolution worse than 2.5. I recommend Dauter Acta Cryst (2003) D59 2004-2016 for a good discussion of this. Your space group might be too high. See the subgroups of P422 at http://cci.lbl.gov/~phzwart/p422_2.png . Reintegrate and merge the data in each space group, MR a single copy of your model (let phaser complete the ASU) and compare the Rpim's (from scaling/merging) and the Rwork/Rfree from a rigid body refine without NCS, with NCS, with appropriate twin laws, and with twin laws + NCS. No need to do a full refinement just yet. Allow phenix.refine to create the Rfree flags. Choose the space group which gives the best statistics. I recently had a case (Hardin, Reyes, Batey J. Biol. Chem., Vol. 284, Issue 22, 15317-15324, May 29, 2009) of a protein that merged into P422 but was difficult to refine in that space group. I brought it back to P4 and refined with NCS+twin to give more reasonable Rwork/ Rfree (5-7% difference from the P422 to P4). HTH, FR On Jul 23, 2009, at 3:54 PM, [email protected] wrote:

Hi Francis, Thanks for your response. The matthews coefficient suggests two molecules in the AU. Phaser also finds two molecules. I ran the dataset through phenix xtriage it did not indicate twinning though. The molecule also exists in nature as a monomer. Shya

...

Twinning? What's your matthews coefficient say? Do you know if your structure is a multimer (biochemistry, etc)? Does it agree with the matthews coefficient?

If the unit cell is not big enough to hold all of the contents,then this is an indicator for twinning .

FR

On Jul 23, 2009, at 3:09 PM, [email protected] wrote:

...

Hi all,

I was trying to solve a structure by molecular replacement. I scaled the data in P4222 space group (resolution 2.7A) with two molecules in the assymmetric unit (molecule A and B) I ran phaser with my model and got a Zscore of 5.1. When I look at the map that I got from phaser I could easily see good electron density for both molecules, However upon inspection of the electron density map there were considerable interaction or clashes with molecule B and a symmetry atom. Molecule A had no clashes however with the symmetry atoms. I was wondering if anyone knows how to resolve this. Could it be a problem of space group. The statistics are good for space group P4222 and the I/sigI was good till 2.7A. Any advice is appreciated, Shya

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

--------------------------------------------- Francis Reyes M.Sc. 215 UCB University of Colorado at Boulder

gpg --keyserver pgp.mit.edu --recv-keys 67BA8D5D

8AE2 F2F4 90F7 9640 28BC 686F 78FD 6669 67BA 8D5D

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

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

--------------------------------------------- Francis Reyes M.Sc. 215 UCB University of Colorado at Boulder gpg --keyserver pgp.mit.edu --recv-keys 67BA8D5D 8AE2 F2F4 90F7 9640 28BC 686F 78FD 6669 67BA 8D5D

Shya I'm curious as to how you resolved this problem. To quote, you applied a twin law for the P 42 22 space group? I wasn't aware that there were twin laws for this space group? How did you determine the twin law? Thanks FR On Jul 24, 2009, at 9:15 AM, [email protected] wrote:

Hi, Thanks for your response. So I did one cycle of refinement and indeed the R work goes down by 5 points when I apply twin law at P4222 space group and when I look at the map the clashes that were present before are no longer there. I will try to scale it in P4 or P422 and see how it looks. I used HKL2000 to scale the data and have no idea if it will make a difference to use SCALA. This is what I got from phenix xtriage: These values look inbetween untwinned and perfect twin Acentric reflections /<I>^2 :1.980 (untwinned: 2.000; perfect twin 1.500) <F>^2/ :0.797 (untwinned: 0.785; perfect twin 0.885) <|E^2 - 1|> :0.727 (untwinned: 0.736; perfect twin 0.541) Centric reflections /<I>^2 :2.863 (untwinned: 3.000; perfect twin 2.000) <F>^2/ :0.673 (untwinned: 0.637; perfect twin 0.785) <|E^2 - 1|> :0.925 (untwinned: 0.968; perfect twin 0.736)

----------------------------------------------- | Z | Nac_obs | Nac_theo | Nc_obs | Nc_theo | ----------------------------------------------- | 0.0 | 0.000 | 0.000 | 0.000 | 0.000 | | 0.1 | 0.074 | 0.095 | 0.208 | 0.248 | | 0.2 | 0.164 | 0.181 | 0.319 | 0.345 | | 0.3 | 0.246 | 0.259 | 0.391 | 0.419 | | 0.4 | 0.322 | 0.330 | 0.452 | 0.474 | | 0.5 | 0.388 | 0.394 | 0.505 | 0.520 | | 0.6 | 0.445 | 0.451 | 0.552 | 0.561 | | 0.7 | 0.497 | 0.503 | 0.592 | 0.597 | | 0.8 | 0.541 | 0.551 | 0.630 | 0.629 | | 0.9 | 0.587 | 0.593 | 0.659 | 0.657 | | 1.0 | 0.631 | 0.632 | 0.679 | 0.683 | ----------------------------------------------- | Maximum deviation acentric : 0.021 | | Maximum deviation centric : 0.040 | | | | _acentric : -0.009 | | _centric : -0.014 |

Thanks again for the valuable input,

Shya

Shya,

...

Did phaser complain that the asymmetric unit was too full? How do the self rotation maps look? Are the crystallographic peaks exact or off by a few degrees (your resolution data may make it difficult to see this)? How do the N(z) cumulative intensity distributions look (make sure to calculate this with thin resolution bins, i.e. increase BINS in Scala I think)? Does your data look sigmoidal on this plot?

Perfect twinning or an NCS that's close to a crystallographic axis is difficult to diagnose from merged intensity statistics and even more difficult with resolution worse than 2.5. I recommend Dauter Acta Cryst (2003) D59 2004-2016 for a good discussion of this.

Your space group might be too high. See the subgroups of P422 at http://cci.lbl.gov/~phzwart/p422_2.png . Reintegrate and merge the data in each space group, MR a single copy of your model (let phaser complete the ASU) and compare the Rpim's (from scaling/merging) and the Rwork/Rfree from a rigid body refine without NCS, with NCS, with appropriate twin laws, and with twin laws + NCS. No need to do a full refinement just yet. Allow phenix.refine to create the Rfree flags. Choose the space group which gives the best statistics.

I recently had a case (Hardin, Reyes, Batey J. Biol. Chem., Vol. 284, Issue 22, 15317-15324, May 29, 2009) of a protein that merged into P422 but was difficult to refine in that space group. I brought it back to P4 and refined with NCS+twin to give more reasonable Rwork/ Rfree (5-7% difference from the P422 to P4).

HTH,

FR

On Jul 23, 2009, at 3:54 PM, [email protected] wrote:

...

Hi Francis, Thanks for your response. The matthews coefficient suggests two molecules in the AU. Phaser also finds two molecules. I ran the dataset through phenix xtriage it did not indicate twinning though. The molecule also exists in nature as a monomer. Shya

...

Twinning? What's your matthews coefficient say? Do you know if your structure is a multimer (biochemistry, etc)? Does it agree with the matthews coefficient?

If the unit cell is not big enough to hold all of the contents,then this is an indicator for twinning .

FR

On Jul 23, 2009, at 3:09 PM, [email protected] wrote:

...

Hi all,

I was trying to solve a structure by molecular replacement. I scaled the data in P4222 space group (resolution 2.7A) with two molecules in the assymmetric unit (molecule A and B) I ran phaser with my model and got a Zscore of 5.1. When I look at the map that I got from phaser I could easily see good electron density for both molecules, However upon inspection of the electron density map there were considerable interaction or clashes with molecule B and a symmetry atom. Molecule A had no clashes however with the symmetry atoms. I was wondering if anyone knows how to resolve this. Could it be a problem of space group. The statistics are good for space group P4222 and the I/sigI was good till 2.7A. Any advice is appreciated, Shya

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

--------------------------------------------- Francis Reyes M.Sc. 215 UCB University of Colorado at Boulder

gpg --keyserver pgp.mit.edu --recv-keys 67BA8D5D

8AE2 F2F4 90F7 9640 28BC 686F 78FD 6669 67BA 8D5D

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

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

--------------------------------------------- Francis Reyes M.Sc. 215 UCB University of Colorado at Boulder

gpg --keyserver pgp.mit.edu --recv-keys 67BA8D5D

8AE2 F2F4 90F7 9640 28BC 686F 78FD 6669 67BA 8D5D

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

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

--------------------------------------------- Francis Reyes M.Sc. 215 UCB University of Colorado at Boulder gpg --keyserver pgp.mit.edu --recv-keys 67BA8D5D 8AE2 F2F4 90F7 9640 28BC 686F 78FD 6669 67BA 8D5D