*** For details on how to be removed from this list visit the *** *** CCP4 home page http://www.ccp4.ac.uk ***
Dear Evette: You need to at least try phaser (which is now integrated with ccp4). (I solved our latest structure with phaser without even having an experimentally derived starting model.) The example in the ccp4 documentation is almost identical to what you describe. There are a lot of testimonials in the ccp4 archives about how phaser allowed people to solve structures that were impossible to solve otherwise, etc. Bill Radisky, Evette S. Ph.D. wrote: > > Dear all, > > In the past I have only ever used epmr for molecular replacement, and > have never encountered any tough cases where that didn't work. I have > never before had to deal with multiple copies in the asymmetric unit. I > am now struggling to solve the structure of a complex between A (a 24 > kDa protein) and B (a 6.5 kDa protein). My native data set extends to > 1.4A resolution; space group is C222 or C2221. For models, I have a > 1.7A structure of A alone, and a 1.6A structure of B alone. Unit cell > analysis suggests that there should be 2 copies of the A-B complex in > the asymmetric unit. (Although the affinity between A and B is not that > great, and it is possible that I have crystallized A alone or B alone; > eg, there could be 3 copies of A and no B in the asymmetric unit.) > > So far, epmr has failed me, and I spent the weekend reading up on Phaser > documentation and trying it out, still with no success. For those > programs, I was trying both possible space groups, and trying to search > with both models as pieces of the complete model, as well as trying to > search with my A model alone, for 2 or 3 copies. > > I am now reading up on Molrep and have started an auto MR run; I guess I > will try the DYAD option next (as mentioned in a recent posting) if the > first run doesn't give a solution. A couple of questions about this: in > using the DYAD=D option for Molrep, how should I choose appropriate > choices for "number of peaks" variables NP, NPT, and NPTD? And in > setting the variables refering to the model, if I am searching for A > alone, is my "fraction completeness of model" 0.8 (80% of one A-B > complex) or 0.4 (40% of the contents of the asymmetric unit)? > > I would also welcome any input about these or other favorite programs > and approaches for situations like this one. Also, I find that program > documentation often doesn't give a lot of guidance for selection of best > options and variables, so any hints on those that may be critical for my > situation would also help. > > Thanks! > Evette > > Evette S. Radisky, Ph.D. > Assistant Professor and Associate Consultant II > Mayo Clinic Cancer Center > Griffin Cancer Research Building, Rm 310 > 4500 San Pablo Road > Jacksonville, FL 32224 > (904) 953-6372 (office) > (904) 953-2857 (lab) > >
