Dear Mahesh, from the images you showed a few days ago I am not convinced the issue is twinning, just overlaps due to the long c-axis. But of course, I do not have have as much info as you, just those images. Unless you are really convinced you have twinning, if you can see what you want to see in the density, perhaps just do not worry too much about it. Or try to collect some better data, with no overlaps, from a crystal mounted with the long axis along the rotation axis, others and us have collected some good data on these kinds of crystals this way. XDS may have integrated your current images reasonably well, but if the crystal is mounted better (mini-kappa goniometer for instance, or a bent loop, or mount many crystals and trust on luck), it will be able to do an even better job. Mark
On 26 Aug 2013, at 17:43, Mahesh Lingaraju wrote: > Hi Juergen & other experts > > Thanks for the suggestions. I was under the impression that the twin > laws/operators are to be used if the twinning is merohedral. In my case, it > appears as if the twinning is non-merohedral and more over the data i have is > processed as P422 which does not have twin operators ( probably because all > the axial pairs are same anyway). Probably because of the same reason, > xtriage did not give any operators for me to use. In such cases, do people > usually try to process the data in another space group of lower symmetry > which has a twin law and is there a definite way to distinguish twinning by > merohedry vs non-merohedry other than just looking at the diffraction pattern > ( is it crucial to make that differentiation ?) ? > > Thanks > > Mahesh > > > On Sun, Aug 25, 2013 at 10:24 PM, Bosch, Juergen <[email protected]> wrote: > Hi Mahesh, > > if you use Refmac, then you can tell it to refine the twin fraction, no need > to tell it the twin law as Refmac will figure it out. If you use phenix, you > explicitly tell it the twin law and refine then with it. You can get the > possible twin laws by running phenix.xtriage and looking at the log file. > > For a 1.7 Å dataset you should see excellent holes in the Phe and Tyr, even > though your Rfactors are high. If that is the case then you are likely > correct with the twin (if nothing else is wrong, Cbeta, Ramas etc). And you > did add some waters to your structure already right ? if not the go water > picking via Coot. > > Have you been converted to XDS now ? Welcome to the club. > > Jürgen > > On Aug 25, 2013, at 8:35 PM, Mahesh Lingaraju wrote: > >> Hello everyone, >> >> I collected a dataset which looked like it is twinned ( or a really long >> axis in the cell) and did not process in HKL2000 and MOSFLM but with some >> of help and suggestions from CCP4BB, XDS was able to process it. The data >> looks good upto 1.7 Å. However, the rfree is stuck at 0.34 even though my >> model is almost complete. I am beginning to wonder if the data is really >> twinned as it has the characteristics of non- merohedral twinning: >> In the images some of the reflections are sharp while some are split and one >> of the axis in the cell is usually long ( the cell is a= 46.78 b= 46.78 c= >> 400.34; 90 90 90) >> >> is there anyway to work around this ? or collecting better data is the only >> solution ? >> >> Any help is deeply appreciated >> >> Thanks >> >> Mahesh > > ...................... > Jürgen Bosch > Johns Hopkins University > Bloomberg School of Public Health > Department of Biochemistry & Molecular Biology > Johns Hopkins Malaria Research Institute > 615 North Wolfe Street, W8708 > Baltimore, MD 21205 > Office: +1-410-614-4742 > Lab: +1-410-614-4894 > Fax: +1-410-955-2926 > http://lupo.jhsph.edu > > > > >
