<div>Yuri,</div>I would refine in parallel with and without the twin law and check the resulting maps and R-factors.<div>I had a case where the twin fraction was 0.06, which I expected to be negotiable. When I included the twin law in refinement, my R-factors dropped dramatically.</div>
<div>It's usually the best rule of thumb that you have the right space group / molecular replacement solution / twinning / etc.</div><div><br></div><div>Good luck,</div><div>Kelly</div><div><br></div><div>*******************************************************<br>
Kelly Daughtry, Ph.D.<br>Post-Doctoral Fellow, Raetz Lab<br>Biochemistry Department<br>Duke University<br>Alex H. Sands, Jr. Building<br>303 Research Drive<br>RM 250<br>Durham, NC 27710<br>P: 919-684-5178<br>*******************************************************<br>
<br><br><div class="gmail_quote">On Thu, Jun 9, 2011 at 3:38 PM, Yuri <span dir="ltr"><<a href="mailto:yuri.pompeu@ufl.edu">yuri.pompeu@ufl.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
Hello phenixers,<br>
I am a little bit at loss since this is the first time I have encountered this situation.<br>
I just collected preliminary data on a ligand soaked crystal. It goes to about 1.85 A. The centrics and acentrics reflections tests look a little deviant from ideal untwinned. I was looking at my xtriage log<br>
file and it seemed to me like I have some twinned fraction. Phenix says not. I realize too my completeness is not good.<br>
Any educated insight would be helpful as I have never dealt with twinning (if this actually the case).<br>
Log file section:<br>
Wilson ratio and moments<br>
<br>
Acentric reflections<br>
<I^2>/<I>^2 :1.650 (untwinned: 2.000; perfect twin 1.500)<br>
<F>^2/<F^2> :0.839 (untwinned: 0.785; perfect twin 0.885)<br>
<|E^2 - 1|> :0.649 (untwinned: 0.736; perfect twin 0.541)<br>
<br>
<br>
Centric reflections<br>
<I^2>/<I>^2 :2.099 (untwinned: 3.000; perfect twin 2.000)<br>
<F>^2/<F^2> :0.749 (untwinned: 0.637; perfect twin 0.785)<br>
<|E^2 - 1|> :0.703 (untwinned: 0.968; perfect twin 0.736)<br>
<br>
<br>
<br>
NZ test (0<=z<1) to detect twinning and possible translational NCS<br>
<br>
<br>
-----------------------------------------------<br>
| Z | Nac_obs | Nac_theo | Nc_obs | Nc_theo |<br>
-----------------------------------------------<br>
| 0.0 | 0.000 | 0.000 | 0.000 | 0.000 |<br>
| 0.1 | 0.052 | 0.095 | 0.171 | 0.248 |<br>
| 0.2 | 0.126 | 0.181 | 0.285 | 0.345 |<br>
| 0.3 | 0.197 | 0.259 | 0.371 | 0.419 |<br>
| 0.4 | 0.271 | 0.330 | 0.440 | 0.474 |<br>
| 0.5 | 0.334 | 0.394 | 0.499 | 0.520 |<br>
| 0.6 | 0.394 | 0.451 | 0.553 | 0.561 |<br>
| 0.7 | 0.447 | 0.503 | 0.600 | 0.597 |<br>
| 0.8 | 0.497 | 0.551 | 0.629 | 0.629 |<br>
| 0.9 | 0.541 | 0.593 | 0.660 | 0.657 |<br>
| 1.0 | 0.581 | 0.632 | 0.697 | 0.683 |<br>
-----------------------------------------------<br>
| Maximum deviation acentric : 0.062 |<br>
| Maximum deviation centric : 0.078 |<br>
| |<br>
| <NZ(obs)-NZ(twinned)>_acentric : -0.050 |<br>
| <NZ(obs)-NZ(twinned)>_centric : -0.021 |<br>
-----------------------------------------------<br>
<br>
<br>
L test for acentric data<br>
<br>
using difference vectors (dh,dk,dl) of the form:<br>
(2hp,2kp,2lp)<br>
where hp, kp, and lp are random signed integers such that<br>
2 <= |dh| + |dk| + |dl| <= 8<br>
<br>
Mean |L| :0.441 (untwinned: 0.500; perfect twin: 0.375)<br>
Mean L^2 :0.270 (untwinned: 0.333; perfect twin: 0.200)<br>
<br>
The distribution of |L| values indicates a twin fraction of<br>
0.00. Note that this estimate is not as reliable as obtained<br>
via a Britton plot or H-test if twin laws are available.<br>
<br>
Cheers,<br><font color="#888888">
-- <br>
Yuri Pompeu<br>
_______________________________________________<br>
phenixbb mailing list<br>
<a href="mailto:phenixbb@phenix-online.org" target="_blank">phenixbb@phenix-online.org</a><br>
<a href="http://phenix-online.org/mailman/listinfo/phenixbb" target="_blank">http://phenix-online.org/mailman/listinfo/phenixbb</a><br>
</font></blockquote></div><br></div>