Dear Appu, You want to be sure you have good reason to drop the space group from C222(1) to P2(1). There may be many reasons why your Rfree may not drop following refinement, especially if you only have one domain in your protein located and just in case there are more molecules to locate in the MR search.
For C222(1) data, did Xtriage suggest any alternate space groups? Also, what program did you use for MR? If you used Phaser with your C222(1) data, did you ask to search for alternate space groups? Did you check for translational NCS? If you want me to take a look at your data, I'd be happy to look at your scaled data and MR results and try to help you out. If so, please email me off the bulletin board. Good luck! Raji On Sun, Mar 24, 2013 at 6:45 AM, vellieux <frederic.velli...@ibs.fr> wrote: > Hello, > > Here we deal with symmetry and the unique part of reciprocal space (the > reciprocal space "asymmetric unit" so to speak). > > C222(1) has eight asymmetric units (international tables, space group 20); > > P2(1) only has two. Assuming that Friedel's law does apply, then the > minimum rotation range to collect a non-redundant data set (one observation > per reflection) is 90 degrees, provided that the crystal is "correctly" and > perfectly aligned. Normally with our current data collection methods where > the crystal is randomly oriented, we would collect more than 90 degrees > (180 degrees, or 360 degrees with the Pilatus detectors on an intense SR > beamline where you cannot really check during data collection how well the > crystal fares during exposure to the X-rays - "shoot first, think later". > > The "reciprocal space" asymmetric unit in C222(1) is smaller. > > I assume that what you are doing is to take the reduced data set file (an > MTZ file probably) and reduce the symmetry from C222(1) to P2(1). You will > not cover the monoclinic reciprocal space asymmetric unit by doing so. > > The way to do it is to take the file from processing, before > (crystallographic symmetry) merging of the equivalents, and perform the > scaling and merging in the P2(1) space group. Or reprocess the data frames > in P2(1) if you have lost the unmerged data file. > > Now of course this will still give you a poor completeness if you have > used a strategy to optimize data collection in the orthorhombic space group > (you won't have collected enough data then for good completeness in the > monoclinic space group). > > I hope this is clear ! > > HTH, > > Fred. > > > On 24/03/13 11:20, Appu kumar wrote: > > I run the phenix.xtriage to evaluate the twining but it suggest no > twining. When i reindex from C2221 to P21, the completeness of data reduced > from 95 % to 35% whereas the map is very good and Rwork and Rfree are 26/31 > for 2.2 resolution. I do not understand why the completeness of data > reduced so much on reindexing. please Can anyone explain this phenomenon. > Thank you > > On 24 March 2013 13:30, Matthias Zebisch < > matthias.zebi...@bbz.uni-leipzig.de> wrote: > >> the p21 c2221 ambivalence can mean severe twinning (i had a similar >> case just now - try several crystals from the same condition) ! >> What do the twinning statistics suggest? >> >> cheers, Matthias >> >> ----------------------------------------- >> Dr. Matthias Zebisch >> Division of Structural Biology, >> Wellcome Trust Centre for Human Genetics, >> University of Oxford, >> Roosevelt Drive, >> Oxford OX3 7BN, UK >> >> Phone (+44) 1865 287549; >> Fax (+44) 1865 287547 >> Email matth...@strubi.ox.ac.uk >> Website http://www.strubi.ox.ac.uk >> ----------------------------------------- >> >> On 3/24/2013 7:46 AM, Appu kumar wrote: >> >> Thank you for the quick reply. After molecular replacement , i have done >> only few cycle of refinement in refmac. I have not done any solvent >> modification or NCS averaging. I have initially indexed the data in C2221 >> but Rfree was not decreasing so i reindexed the data in data in P121 space >> group keeping the Rfree flag of C2221. While analysing the symmetry mates , >> i found large space but no density. structure of Ligand binding domain is >> almost identical with 90% identity in sequence. I am stuck with this >> problem and don't know how to process further. >> Please give me your valuable suggestion. I will appreciate your effort. >> Thank you >> Appu >> >> On 24 March 2013 02:38, Raji Edayathumangalam <r...@brandeis.edu> wrote: >> >>> Dear Appu, >>> >>> I am not sure that I have a complete sense of the issue at hand since >>> some of the information needed to think your issue through is missing in >>> your email. For example, to what high resolution cut-off were the data >>> measured? What resolution limits were used for the MR search? How do the >>> unit cell dimensions and space group in the two cases compare? >>> >>> I am guessing the ligand binding domain in your protein has the >>> identical sequence to that of the published ligand binding domain that you >>> use as a template in your MR search. In any case, here are a couple of my >>> thoughts: >>> >>> (1) It might be worth setting up different runs of MR with different >>> numbers for expected copies (not just two copies but also one copy and >>> three copies just in case you have one of the extreme cases of solvent >>> content)? >>> >>> (2) If the MR solution is correct and there is physical room for a DNA >>> binding domain in your lattice (check by displaying symmetry mates), >>> perhaps the DNA binding domain is disordered. In that case (and if all >>> attempts with current data fail), you may have to crystallize the protein >>> in presence of DNA. >>> >>> >>> Good luck! >>> Raji >>> >>> >>> >>> >>> On Sat, Mar 23, 2013 at 2:26 PM, Appu kumar <appu.kum...@gmail.com>wrote: >>> >>>> Dear members, >>>> >>>> I am doing a molecular replacement of a >>>> transcription factor whose ligand binding structure(24000 Da) is available >>>> in PDB but not for the DNA binding(13000 Da). When i am searching for the >>>> two copies from ligand binding domain as a template model, i am getting >>>> very good solution but i am not getting any density for the DNA binding >>>> domain to build up in density. The space gorup is P 1 21 1 (4) and unit >>>> cell parameters are Unit Cell: 57.43 69.36 105.99 90.00 90.00 >>>> 90.00. Please guide me how to get the complete model structure. Table below >>>> show the matthews statistics >>>> >>>> For estimated molecular weight 37000. >>>> Nmol/asym Matthews Coeff %solvent P(2.20) P(tot) >>>> _____________________________________________________________ >>>> 1 5.71 78.46 0.00 0.01 >>>> 2 2.85 56.91 0.62 0.70 >>>> 3 1.90 35.37 0.37 0.29 >>>> 4 1.43 13.82 0.00 0.00 >>>> _____________________________________________________________ >>>> >>>> >>>> The phaser molecular replacement gives the following table. >>>> istogram of relative frequencies of VM values >>>> ---------------------------------------------- >>>> Frequency of most common VM value normalized to 1 >>>> VM values plotted in increments of 1/VM (0.02) >>>> >>>> <--- relative frequency ---> >>>> 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 >>>> | | | | | | | | | | | >>>> 10.00 - >>>> 8.33 - >>>> 7.14 - >>>> 6.25 - >>>> 5.56 - >>>> 5.00 - >>>> 4.55 - >>>> 4.17 - >>>> 3.85 -- >>>> 3.57 --- >>>> 3.33 ------ >>>> 3.12 ---------- >>>> 2.94 **************** (COMPOSITION*1) >>>> 2.78 ----------------------- >>>> 2.63 -------------------------------- >>>> 2.50 ----------------------------------------- >>>> 2.38 ------------------------------------------------ >>>> 2.27 -------------------------------------------------- >>>> 2.17 ----------------------------------------------- >>>> 2.08 -------------------------------------- >>>> 2.00 -------------------------- >>>> 1.92 --------------- >>>> 1.85 ------- >>>> 1.79 --- >>>> 1.72 - >>>> 1.67 - >>>> 1.61 - >>>> 1.56 - >>>> 1.52 - >>>> 1.47 * (COMPOSITION*2) >>>> 1.43 - >>>> 1.39 - >>>> 1.35 - >>>> 1.32 - >>>> 1.28 - >>>> 1.25 - >>>> >>>> $TABLE : Cell Content Analysis: >>>> $SCATTER >>>> :N*Composition vs Probability:0|3x0|1:1,2: >>>> $$ >>>> N*Composition Probability >>>> $$ loggraph $$ >>>> 1 0.306066 >>>> 2 0.00141804 >>>> $$ >>>> >>>> Most probable VM for resolution = 2.27817 >>>> Most probable MW of protein in asu for resolution = 92664.2 >>>> >>>> Thank a lot in advance >>>> >>>> >>>> >>>> >>> >>> -- >>> Raji Edayathumangalam >>> Instructor in Neurology, Harvard Medical School >>> Research Associate, Brigham and Women's Hospital >>> Visiting Research Scholar, Brandeis University >>> >>> >> >> > > > -- > Fred. Vellieux (B.Sc., Ph.D., hdr) > ouvrier de la recherche > IBS / ELMA > 41 rue Jules Horowitz > F-38027 Grenoble Cedex 01 > Tel: +33 438789605 > Fax: +33 438785494 > > -- Raji Edayathumangalam Instructor in Neurology, Harvard Medical School Research Associate, Brigham and Women's Hospital Visiting Research Scholar, Brandeis University
