You might find Structure 14, 973-982 Jun 2006 of interest. Larger protein, Zn instead of Se (although only 8 sites), roughly comparable to slightly lower resolution. This was mainly testing to confirm that there was enough anomalous signal to phase off of, so the model was essentially known; however 50% completeness was enough to allow for site location by anomalous or dispersive difference fourier (which one sticking point...even if the MR hits don't give a helpful model phased map, they can still be good enough for site location).
Pete Clemens Grimm wrote: > OK, here's a concrete case: > > A 150kDa protein complex, the plate-like crystals can be produced in > sufficient number; Se-Met derivatives available, total number of Met > around 20, subunits could be marked and combined individually. > Diffraction is highly anisotropic, in certain directions up to > 3.8A,while in others only 5A. Similarly, the spot quality is very > dependent on orientation. > Space group I222, a=75 b=150 c=250. Datsets scale well with 3-4% Rsym > up to 12 A resolution. At 4.5A Rsym rises above 50% (I/sigma is still at > 2.0). The 'sweet' slices of the dataset scale significantly better, but > give only 70% (non-anomalous) completness. We hope to improve datasets > slightly by orienting the crystals. > 65% of the structure would be available as coordinate building blocks > from the PDB, however, MR with these components so far did not yield a > clear solution. > > Any suggestions or experiences with similar cases are welcome. > > Cheers, > Clemens > > Zitat von Clemens Vonrhein <vonrh...@globalphasing.com>: > > > [Show Quoted Text - 64 lines][Zitattext verstecken] > Hi Clemens, > > maybe re-phrasing your question: > > What would be the best technique/strategy to phase crystals that > > [ ] diffract to maximal ___ A > [ ] typical diffraction to __ A > [ ] are radiation sensitive > [ ] easily reproducable > [ ] large crystals (up to ___ um) > [ ] long needles > [ ] thin plates > [ ] have ___ mol/asu > [ ] spacegroup ___ > [ ] nice diffraction pattern > [ ] poor diffraction pattern (reason: ___) > [ ] anisotropic diffraction (resolution in poorest direction: ___ A) > [ ] cell dimensions of roughly ___ ___ ___ ___ ___ ___ > [ ] purified from native source > [ ] expressed in expression system ___ > [ ] anything else: _______ > > Tick the appropriate boxes and fill out the blanks as much as possible > - that should give more important and necessary information. There are > consequences to consider for all of those points that would then give > some rough guidelines for your particular project/problem. > > Maybe CCP4 should have an online form to describe a particular > crystallographic problem? > > Cheers > > Clemens > > On Thu, May 14, 2009 at 09:35:28AM +0200, Clemens Grimm wrote: > > Dear all, > > after the SeMet phasing discussion, what would be -in general- the best > technique to phase low resolution data (<=4A) of large complexes (>=150 > kDA) - in terms of > > - derivatization compounds (is there something like the 'golden five' HA > compounds for these cases), > > - data collection techniques and > > - phasing methods? > > Clemens > > -- > > *************************************************************** > * Clemens Vonrhein, Ph.D. vonrhein AT GlobalPhasing DOT com > * > * Global Phasing Ltd. > * Sheraton House, Castle Park > * Cambridge CB3 0AX, UK > *-------------------------------------------------------------- > * BUSTER Development Group (http://www.globalphasing.com) > ***************************************************************
