Matthew, Here's my $0.02:
Try more sophisticated twin tests such as xtriage from phenix or Ctruncate from the latest CCP4 distro. Both programs include multiple twining tests. Second, your R-factors seem very reasonable for a well refined structure missing 20% of a not-so-well ordered domain. If it ain't broke? Thirdly, if you do have a true twin fraction of 0.5, you have a hemihedral twin which cannot be detwinned -- I believe much of the "detwinned" data results in F = 0, could be wrong about that. Also, is your data anisotropic (your cell constants seem to be)? Anisotropy can interfere with twinning analysis. Lastly, think about optimizing your solvent envelope - this can make weak density stronger and improve R-factors as well (the latest CNS has a good implementation of optimized solvent flattening). Best of luck, --Paul --- On Wed, 7/22/09, Matthew Franklin <matthew.frank...@imclone.com> wrote: > From: Matthew Franklin <matthew.frank...@imclone.com> > Subject: [ccp4bb] is my crystal twinned or not? > To: CCP4BB@JISCMAIL.AC.UK > Date: Wednesday, July 22, 2009, 6:02 PM > Hi all - > > I'm trying to solve the structure of an antibody-receptor > complex, and I've hit a wall which may be due to a twinned > crystal form. This crystal has the apparent space > group P43212, with cell constants a=64.02 c=274.83. > Solvent content analysis using this space group suggests 1 > mol/asu, with 47% solvent, which would be fairly consistent > with the diffraction limit of about 2.8 A. I've been > able to place the antibody using molecular replacement and > refine it to R=0.289, Rf=0.338. There is some density > for the receptor, which represents about 20% of the mass of > the complex, but not clear enough to build into, and all > efforts to improve the density or place the receptor by > molecular replacement have failed. > > Well, tough luck, you might say, but I noticed at the very > beginning of the process that this crystal form may be a > perfect twin. The "4th moment of E" graph from > Truncate shows nearly all resolution bins with values of 1.4 > - 1.6, except in the very topmost few bins where the values > rise up to 2. The other moment graphs are likewise at > their "perfect twin" values across the resolution > range. The cumulative intensity distribution graph > shows the observed values are about 30% lower than the > expected values for both acentric and centric > reflections. As I understand it, both of these are > strong indicators of twinning, and the twin fraction > analysis in DETWIN suggests a twin fraction of ~0.5. > > So what do I do now? I think I'm supposed to > reprocess the data in the space group without the twin > transformation (which would be P43), then run molecular > replacement which should show me solutions for both halves > of the twin. However, all I'm seeing is two copies of > the antibody in the (P43) asymmetric unit, which don't > overlap, and both of which need to be present in the same > unit cell in order to form a 3-dimensional lattice. > This is exactly what I would expect to see if there were no > twinning present. > > So my question is, is this crystal form twinned or > not? Is there some way that the intensity statistics > could be misleading me? On the other side, am I doing > something wrong with the structure determination if the > crystal is twinned? How should I proceed? > > Thanks for any help anyone can provide. > > - Matt > > > -- > Matthew Franklin , Ph.D. > Senior Scientist, ImClone Systems, > a wholly owned subsidiary of Eli Lilly & Company > 180 Varick Street, 6th floor > New York, NY 10014 > phone:(917)606-4116 fax:(212)645-2054 > > > > Confidentiality Note: > This e-mail, and any attachment to it, contains privileged > and confidential information intended only for the use of > the individual(s) or entity named on the e-mail. If the > reader of this e-mail is not the intended recipient, or the > employee or agent responsible for delivering it to the > intended recipient, you are hereby notified that reading it > is strictly prohibited. If you have received this e-mail in > error, please immediately return it to the sender and delete > it from your system. > > Thank you. >
