[ccp4bb] Twin refinement in Refmac
Dear colleagues, I am in the process of refining a P2(1) structure that is marked by Scala as twinned (L-test etc.) with a twinning fraction of ca. 0.2. I am still working on the twinning problem, so no more details about this. But what puzzles me right now is that, if I feed the data with a max. resolution of 1.95 A into Refmac with the twin refinement option on, it recognizes the resolution cut-off at first, but then uses a new cut-off of 1.88 A in the first ( and all subsequent) refinement cycles. Since there are no data at this resolution, this seems to screw up the calculation of R-values at high resolution and it reports for them zero-values in the co-ordinate header. Without twin refinement, the max. resolution stays at 1.95 A (but the final stats are worse, of course). Is this a bug or am I missing here something? Thanks in advance, and I am happy to send log files or my precious data if necessary. Cheers, Uli --- dr ulrich gohlke staff scientist - macromolecular structure and interaction max-delbrück-center for molecular medicine (mdc) +49 30 9406 - 2725 (w) +49 30 9406 - 2548 (fax) ulrich.goh...@mdc-berlin.demailto:ulrich.goh...@mdc-berlin.de http://www.mdc-berlin.de/en/research/research_teams/macromolecular_structure_and_interaction/
Re: [ccp4bb] Twin refinement in Refmac
Dear Uli, It seems that you are dealing with psued-twinning when cell do not overlap after rotation with twin operator. In these case what is in one resolution on one of the orientation becomes another resolution in another orientation of crystals In short:. You have an operator R that relates two orientation of the crystal. If R does not convert cell to itself completely then resolution of twin related reflections h' and h h' = Rh will be different. It usually means that in your images spots are split at higher resolution. If you have not integrated those spots as one at higher resolution then refinement may not be optimal. These cases needs to be considered as multiple orientation of the same crystal. I think MOSFLM should be able to integrate them as several cells. However refmac is not completely ready to deal with these cases as optimally as it should. I hope it clears things a bit. Cheers, Garib On 22 May 2014, at 15:48, ulrich.goh...@mdc-berlin.de ulrich.goh...@mdc-berlin.de wrote: Dear colleagues, I am in the process of refining a P2(1) structure that is marked by Scala as twinned (L-test etc.) with a twinning fraction of ca. 0.2. I am still working on the twinning problem, so no more details about this. But what puzzles me right now is that, if I feed the data with a max. resolution of 1.95 A into Refmac with the twin refinement option on, it recognizes the resolution cut-off at first, but then uses a new cut-off of 1.88 A in the first ( and all subsequent) refinement cycles. Since there are no data at this resolution, this seems to screw up the calculation of R-values at high resolution and it reports for them zero-values in the co-ordinate header. Without twin refinement, the max. resolution stays at 1.95 A (but the final stats are worse, of course). Is this a bug or am I missing here something? Thanks in advance, and I am happy to send log files or my precious data if necessary. Cheers, Uli --- dr ulrich gohlke staff scientist - macromolecular structure and interaction max-delbrück-center for molecular medicine (mdc) +49 30 9406 - 2725 (w) +49 30 9406 - 2548 (fax) ulrich.goh...@mdc-berlin.de http://www.mdc-berlin.de/en/research/research_teams/macromolecular_structure_and_interaction/ Dr Garib N Murshudov MRC-LMB Francis Crick Avenue Cambridge CB2 0QH UK Web http://www.mrc-lmb.cam.ac.uk, http://www2.mrc-lmb.cam.ac.uk/groups/murshudov/
Re: [ccp4bb] Twin refinement in Refmac
Dear Garib, thanks for the clarification. I'll go back to the raw data and check for split spots at high resolution (and give Mosflm a try; initially, the data were integrated with XDS). Cheers, Uli --- dr ulrich gohlke staff scientist - macromolecular structure and interaction max-delbrück-center for molecular medicine (mdc) +49 30 9406 - 2725 (w) +49 30 9406 - 2548 (fax) ulrich.goh...@mdc-berlin.de http://www.mdc-berlin.de/en/research/research_teams/macromolecular_structure_and_interaction/
Re: [ccp4bb] twin refinement
If the twin law is k,h,-l, then your a axis must almost equal the b axis? And if the twin fraction is 0.48 then you have additional symmetry I guess? How sure are you that the point group is P4/mmm? On 13 March 2014 20:41, Teresa Swanson teresa.m.swan...@gmail.com wrote: Dear collegues, I'm working with a drug complexed protein structure that is having major twinning issues. The drug has a single Br atom on a benzene ring, which I'd like to use for orienting the drug in the binding site. I have various anomalous data sets, ranging from 3.0A resolution, all scaled into P222 with a Rlin of .125. Using MR, the twin law (k, h, -l) and NCS restraints, I can confidently solve the structure without anomalous, and the drug density is clear in the Fo-Fc map, with Rw/Rf at ~.26/.29 and a space group of P21221. It might be important to note that any simulated annealing I've tried invariably increases the Rfree by 2-3%, so I've scraped it. As you can imagine, when using the twinned data, the anomalous maps are weak and random. I've used the Phenix detwin option in Xtriage to see if I can pull the anomalous signal out of it. If I use the .mtz file that is output for MR and calculate the anomalous maps, it looks promising. The twin fraction for the one particular dataset I've been using is estimated at approx .48. Is this too close to 50% to do the detwinning? Now I'm wondering how to properly refine this further. I'm assuming that since I've detwinned the data, I do refinement without the twin law. But that gives an initial Rf of .38 when using it gives .31. Since I've already solved the structure without using the anomalous flag, can I just use the detwinned reflections and the refined structure to calculate an anomalous map (without having to redo the refinement)? Mainly, my main question is about how to tease out and properly refine the anomalous data from a twinned structure. Also, how much of a difference will it make to scale into P222 versus P21212. And, if I have quite high redundancy, should I scale anomalous in HKL2000 or just use the anomalous flag? Any help on refining this twinned structure would be greatly appreciated! Thanks, Teresa PhD Student
Re: [ccp4bb] twin refinement
Hi Teresa, As Eleanor has mentioned, you should probably check out other space groups. Xtriage gives a lot of great information and many plots to inspect. But, if you do not know what the plots mean and just look at the results that say the twin fraction is 0.48 you can get into some trouble. From what I have seen over the years, when most people say they have a twin fraction nearly 0.5, they are usually in the wrong space group and the data is not twinned. When the summary table shows a low 'R obs' and a high twin fraction across the tests that means that the operator is present in the data. But that could be a crystallographic operator or a twin operator. You would have to inspect the plots to see if the data looks twinned, but if there is pseudosymmetry it could make the plots look not twinned. You could run the extra twin tests in Xtriage inputting the MR solution so that the R vs. R plot (Andrey Lebedev, et al. (2006) , Acta Crystallogr. D62, 83-95) could be calculated. In the end, your data could be nearly perfectly twinned, I have had it happen a couple of times, but I always tell people to assume it is not until the model building is done and the R-factor still won't drop. Your R-factors are not that bad for 3A resolution, indicating the operator might be crystallographic and you are too low in symmetry. So if you have done all this and you still believe the data is twinned... Don't detwin your data! Your twin fraction is too high and it will introduce a lot of errors. This has been gone over many times and you can search the archives. I never detwin data regardless of twin fraction. Modern programs like Phenix.refine and Refmac (I think) handle twinned data just fine. If you are just trying to calculate an ADF map to find the Br then it might be difficult since the peak heights are going to be lower. Your best bet is to increase your redundancy without killing the crystal with too much radiation damage. If your redundancy is around 10 or so then you could probably use the 'scale anomalous' option. I tend to scale with and without it and look at the maps to see if it helps or not. Sometimes it helps, other times it doesn't and I am sure there are reasons (like anisotropy). Another word of caution is the possibility of model bias in the maps when the twin fraction approaches 0.5. The programs will use the model in the detwinning process if the twin fraction is above a certain threshold. I believe it is 0.45 in Phenix.refine. First thing I would do is follow Eleanor's lead and check out other space groups. You might have pseudo-symmetry with or without twinning. Jon -- Jonathan P. Schuermann, Ph. D. Beamline Scientist, NE-CAT Argonne National Laboratory, 436E 9700 S. Cass Ave. Argonne, IL 60439 Email: schue...@anl.gov Tel: (630) 252-0682 On 03/13/2014 03:41 PM, Teresa Swanson wrote: Dear collegues, I'm working with a drug complexed protein structure that is having major twinning issues. The drug has a single Br atom on a benzene ring, which I'd like to use for orienting the drug in the binding site. I have various anomalous data sets, ranging from 3.0A resolution, all scaled into P222 with a Rlin of .125. Using MR, the twin law (k, h, -l) and NCS restraints, I can confidently solve the structure without anomalous, and the drug density is clear in the Fo-Fc map, with Rw/Rf at ~.26/.29 and a space group of P21221. It might be important to note that any simulated annealing I've tried invariably increases the Rfree by 2-3%, so I've scraped it. As you can imagine, when using the twinned data, the anomalous maps are weak and random. I've used the Phenix detwin option in Xtriage to see if I can pull the anomalous signal out of it. If I use the .mtz file that is output for MR and calculate the anomalous maps, it looks promising. The twin fraction for the one particular dataset I've been using is estimated at approx .48. Is this too close to 50% to do the detwinning? Now I'm wondering how to properly refine this further. I'm assuming that since I've detwinned the data, I do refinement without the twin law. But that gives an initial Rf of .38 when using it gives .31. Since I've already solved the structure without using the anomalous flag, can I just use the detwinned reflections and the refined structure to calculate an anomalous map (without having to redo the refinement)? Mainly, my main question is about how to tease out and properly refine the anomalous data from a twinned structure. Also, how much of a difference will it make to scale into P222 versus P21212. And, if I have quite high redundancy, should I scale anomalous in HKL2000 or just use the anomalous flag? Any help on refining this twinned structure would be greatly appreciated! Thanks, Teresa PhD Student -- Jonathan P. Schuermann, Ph. D. Beamline Scientist, NE-CAT Argonne National Laboratory, 436E 9700 S. Cass Ave. Argonne, IL 60439 Email: schue...@anl.gov Tel:
[ccp4bb] twin refinement
Dear colleagues, this is a request for comments on the evaluation of crystal structures that resulted from twin refinement. From http://www.ysbl.york.ac.uk/~garib/refmac/Tutorials/refmac_tutorial.pdf: Although Rfactors are substantially smaller with twin refinement than without twin refinement it does not mean that model also is substantially better. [...] Usually twin maps after twin refinement are more biased towards errors in the model. From phenix.xtriage output: Please note however that R-factors from twinned refinement cannot be directly compared to R-factors without twinning, as they will always be lower when a twin law is used. You should also use caution when interpreting the maps from refinement, as they will have significantly more model bias. How would you ascertain that inclusion of twin parameters has improved the model? How is one to judge the evidentiary strength of such crystal structure in general, given that validation by residuals and maps is further weakened compared to refinement without twinning parameters? Thank you for your thoughts and time, Wolfram Tempel
Re: [ccp4bb] twin refinement
Dear Wolfram, First of all you should make sure you have substantial twinning. Is the twin fraction close to zero, then don't use a twin target. Is twinning detected by several test, then there is a good chance your data are twinned. As for validation, if you do not completely trust your R-factors or maps, spend more time looking at geometry. That is independent of your diffraction data, but still a very good way to see whether your model is plausible. Hth, Robbie Sent from my Windows Phone Van: wtempel Verzonden: 13-3-2014 15:19 Aan: CCP4BB@JISCMAIL.AC.UK Onderwerp: [ccp4bb] twin refinement Dear colleagues, this is a request for comments on the evaluation of crystal structures that resulted from twin refinement. From http://www.ysbl.york.ac.uk/~garib/refmac/Tutorials/refmac_tutorial.pdf: Although Rfactors are substantially smaller with twin refinement than without twin refinement it does not mean that model also is substantially better. [...] Usually twin maps after twin refinement are more biased towards errors in the model. From phenix.xtriage output: Please note however that R-factors from twinned refinement cannot be directly compared to R-factors without twinning, as they will always be lower when a twin law is used. You should also use caution when interpreting the maps from refinement, as they will have significantly more model bias. How would you ascertain that inclusion of twin parameters has improved the model? How is one to judge the evidentiary strength of such crystal structure in general, given that validation by residuals and maps is further weakened compared to refinement without twinning parameters? Thank you for your thoughts and time, Wolfram Tempel
[ccp4bb] twin refinement
Dear collegues, I'm working with a drug complexed protein structure that is having major twinning issues. The drug has a single Br atom on a benzene ring, which I'd like to use for orienting the drug in the binding site. I have various anomalous data sets, ranging from 3.0A resolution, all scaled into P222 with a Rlin of .125. Using MR, the twin law (k, h, -l) and NCS restraints, I can confidently solve the structure without anomalous, and the drug density is clear in the Fo-Fc map, with Rw/Rf at ~.26/.29 and a space group of P21221. It might be important to note that any simulated annealing I've tried invariably increases the Rfree by 2-3%, so I've scraped it. As you can imagine, when using the twinned data, the anomalous maps are weak and random. I've used the Phenix detwin option in Xtriage to see if I can pull the anomalous signal out of it. If I use the .mtz file that is output for MR and calculate the anomalous maps, it looks promising. The twin fraction for the one particular dataset I've been using is estimated at approx .48. Is this too close to 50% to do the detwinning? Now I'm wondering how to properly refine this further. I'm assuming that since I've detwinned the data, I do refinement without the twin law. But that gives an initial Rf of .38 when using it gives .31. Since I've already solved the structure without using the anomalous flag, can I just use the detwinned reflections and the refined structure to calculate an anomalous map (without having to redo the refinement)? Mainly, my main question is about how to tease out and properly refine the anomalous data from a twinned structure. Also, how much of a difference will it make to scale into P222 versus P21212. And, if I have quite high redundancy, should I scale anomalous in HKL2000 or just use the anomalous flag? Any help on refining this twinned structure would be greatly appreciated! Thanks, Teresa PhD Student
Re: [ccp4bb] twin refinement in refmac
Is this twinning or several crystals indexed according to different conventions? You usually see evidence of twinning for each crystal if it is really there.. Trigonal data can be indexed as h,k,l k,h,-k -h,-k,l or -k,-h,-l of course so you have a 75% chance of getting the 2nd crystal on a different convention than the first.. POINTLESS checks for this if you give it pairs of input files and does a good job in selecting the same indexing convention - providing of course the twinning isnt present. I would be suspicious because a) I have been convinced by having more than 2 twin domains b) when there is twinning the data analysis on each crystal seperately shows it up. Eleanor - On Mar 2 2012, wtempel wrote: Dear CCp4ers, A good morning to everyone. Today, I have a structure that I initially refined in space group P6522, 1mol/asu. Scaling stats (scalepack): 2.30-2.26A: Rsym=99.9%; I/sigma 3 2.61-2.55A: Rsym=39.6%, I/sigma 10 50.00-6.13: Rsym=6.4% Some mild anisotropy in the resolution limits is apparent on the diffraction images. Say, visible spots at 2.2A in one direction, 2.6A in the other. Rfree, using data to 2.3A, was stuck in mid-30%s. The map appears like 3.5A resolution, with some difference density for loops that cannot be interpreted with reasonable geometry. Rsym is very similar for data scaled in P3, in all resolution shells. Xtriage does not suggest merohedral twinning. Nevertheless, I extended my free flags in sftools from P6522 to P32 and cad'd them to amplitudes merged in spacegroup P32. Correspondingly, I expanded my model to a homotetramer and ran Refmac with amplitude based twinning. (Would this be a reasonable input to twin refinement?) From the output coordinates: REMARK 3 TWIN DETAILS REMARK 3 NUMBER OF TWIN DOMAINS :4 REMARK 3 TWIN DOMAIN :1 REMARK 3 TWIN OPERATOR : H, K, L REMARK 3 TWIN FRACTION : 0.269 REMARK 3 TWIN DOMAIN :2 REMARK 3 TWIN OPERATOR : -K, -H, -L REMARK 3 TWIN FRACTION : 0.171 REMARK 3 TWIN DOMAIN :3 REMARK 3 TWIN OPERATOR : K, H, -L REMARK 3 TWIN FRACTION : 0.258 REMARK 3 TWIN DOMAIN :4 REMARK 3 TWIN OPERATOR : -H, -K, L REMARK 3 TWIN FRACTION : 0.302 Does this establish twinning versus underestimated symmetry? And what do I need to know about my free-R? Did refmac assign a new flag? Whereas the output file's flags are all 1s and 0s, the input file had 0 ... 19. During the first run, Rfree dropped to 28%. But on a subsequent run, Rfree was stuck 30% when I used the initial job's output MTZ. Many thanks in advance for your helpful comments. Wolfram Tempel -- Professor Eleanor Dodson YSNL, Dept of Chemistry University of York Heslington YO10 5YW tel: 00 44 1904 328259 Fax: 00 44 1904 328266
Re: [ccp4bb] twin refinement in refmac
Hi all, I will like to know, as a follow up of what Prof. Randy Read said, what should be done to do the refinement against the measured data and not the detwinned F( which refmac outputs in the mtz after twin refinement), during subsequent refinements. And also, I would like to know how to ensure that the free R generated takes twinning into account if I am not using phenix. Thanks in advance, Regards, ARKO On Fri, Mar 2, 2012 at 12:44 PM, Randy J. Read rj...@cam.ac.uk wrote: I'm worried when you say that you use the initial job's output MTZ. Refmac replaces F with a detwinned F in the output file so you wouldn't be refining against your measured data in the subsequent round. Best wishes Randy Read Randy J. Read On 2 Mar 2012, at 02:00, wtempel wtem...@gmail.com wrote: Dear CCp4ers, A good morning to everyone. Today, I have a structure that I initially refined in space group P6522, 1mol/asu. Scaling stats (scalepack): 2.30-2.26A: Rsym=99.9%; I/sigma 3 2.61-2.55A: Rsym=39.6%, I/sigma 10 50.00-6.13: Rsym=6.4% Some mild anisotropy in the resolution limits is apparent on the diffraction images. Say, visible spots at 2.2A in one direction, 2.6A in the other. Rfree, using data to 2.3A, was stuck in mid-30%s. The map appears like 3.5A resolution, with some difference density for loops that cannot be interpreted with reasonable geometry. Rsym is very similar for data scaled in P3, in all resolution shells. Xtriage does not suggest merohedral twinning. Nevertheless, I extended my free flags in sftools from P6522 to P32 and cad'd them to amplitudes merged in spacegroup P32. Correspondingly, I expanded my model to a homotetramer and ran Refmac with amplitude based twinning. (Would this be a reasonable input to twin refinement?) From the output coordinates: REMARK 3 TWIN DETAILS REMARK 3 NUMBER OF TWIN DOMAINS :4 REMARK 3 TWIN DOMAIN :1 REMARK 3 TWIN OPERATOR : H, K, L REMARK 3 TWIN FRACTION : 0.269 REMARK 3 TWIN DOMAIN :2 REMARK 3 TWIN OPERATOR : -K, -H, -L REMARK 3 TWIN FRACTION : 0.171 REMARK 3 TWIN DOMAIN :3 REMARK 3 TWIN OPERATOR : K, H, -L REMARK 3 TWIN FRACTION : 0.258 REMARK 3 TWIN DOMAIN :4 REMARK 3 TWIN OPERATOR : -H, -K, L REMARK 3 TWIN FRACTION : 0.302 Does this establish twinning versus underestimated symmetry? And what do I need to know about my free-R? Did refmac assign a new flag? Whereas the output file's flags are all 1s and 0s, the input file had 0 ... 19. During the first run, Rfree dropped to 28%. But on a subsequent run, Rfree was stuck 30% when I used the initial job's output MTZ. Many thanks in advance for your helpful comments. Wolfram Tempel -- *ARKA CHAKRABORTY* *CAS in Crystallography and Biophysics* *University of Madras* *Chennai,India*
Re: [ccp4bb] twin refinement in refmac
Garib may have more to say, but the first point would be to always include the original data file as your input MTZ file for any cycle of refinement, whether you're using Refmac in CCP4 or phenix.refine. (In phenix.refine, if you assign the R-free data the first time you do refinement, it will produce a file containing all the information used in that cycle of refinement, including the new R-free flags and possibly any Hendrickson-Lattman coefficients, and you should use that file for all subsequent refinements.) As for whether R-free takes twinning into account, I think it's fair to say that all the refinement programs that handle twinning will do this appropriately. Regards, Randy Read On 2 Mar 2012, at 08:01, arka chakraborty wrote: Hi all, I will like to know, as a follow up of what Prof. Randy Read said, what should be done to do the refinement against the measured data and not the detwinned F( which refmac outputs in the mtz after twin refinement), during subsequent refinements. And also, I would like to know how to ensure that the free R generated takes twinning into account if I am not using phenix. Thanks in advance, Regards, ARKO On Fri, Mar 2, 2012 at 12:44 PM, Randy J. Read rj...@cam.ac.uk wrote: I'm worried when you say that you use the initial job's output MTZ. Refmac replaces F with a detwinned F in the output file so you wouldn't be refining against your measured data in the subsequent round. Best wishes Randy Read Randy J. Read On 2 Mar 2012, at 02:00, wtempel wtem...@gmail.com wrote: Dear CCp4ers, A good morning to everyone. Today, I have a structure that I initially refined in space group P6522, 1mol/asu. Scaling stats (scalepack): 2.30-2.26A: Rsym=99.9%; I/sigma 3 2.61-2.55A: Rsym=39.6%, I/sigma 10 50.00-6.13: Rsym=6.4% Some mild anisotropy in the resolution limits is apparent on the diffraction images. Say, visible spots at 2.2A in one direction, 2.6A in the other. Rfree, using data to 2.3A, was stuck in mid-30%s. The map appears like 3.5A resolution, with some difference density for loops that cannot be interpreted with reasonable geometry. Rsym is very similar for data scaled in P3, in all resolution shells. Xtriage does not suggest merohedral twinning. Nevertheless, I extended my free flags in sftools from P6522 to P32 and cad'd them to amplitudes merged in spacegroup P32. Correspondingly, I expanded my model to a homotetramer and ran Refmac with amplitude based twinning. (Would this be a reasonable input to twin refinement?) From the output coordinates: REMARK 3 TWIN DETAILS REMARK 3 NUMBER OF TWIN DOMAINS :4 REMARK 3 TWIN DOMAIN :1 REMARK 3 TWIN OPERATOR : H, K, L REMARK 3 TWIN FRACTION : 0.269 REMARK 3 TWIN DOMAIN :2 REMARK 3 TWIN OPERATOR : -K, -H, -L REMARK 3 TWIN FRACTION : 0.171 REMARK 3 TWIN DOMAIN :3 REMARK 3 TWIN OPERATOR : K, H, -L REMARK 3 TWIN FRACTION : 0.258 REMARK 3 TWIN DOMAIN :4 REMARK 3 TWIN OPERATOR : -H, -K, L REMARK 3 TWIN FRACTION : 0.302 Does this establish twinning versus underestimated symmetry? And what do I need to know about my free-R? Did refmac assign a new flag? Whereas the output file's flags are all 1s and 0s, the input file had 0 ... 19. During the first run, Rfree dropped to 28%. But on a subsequent run, Rfree was stuck 30% when I used the initial job's output MTZ. Many thanks in advance for your helpful comments. Wolfram Tempel -- ARKA CHAKRABORTY CAS in Crystallography and Biophysics University of Madras Chennai,India -- Randy J. Read Department of Haematology, University of Cambridge Cambridge Institute for Medical Research Tel: + 44 1223 336500 Wellcome Trust/MRC Building Fax: + 44 1223 336827 Hills RoadE-mail: rj...@cam.ac.uk Cambridge CB2 0XY, U.K. www-structmed.cimr.cam.ac.uk
Re: [ccp4bb] twin refinement in refmac
1) You should use measured data (after scala/aimless/truncate). In general there may not be one to one relationship between observed data and asymmetric unit (e.g. non-merohedral twinning) and it would not be possible to bring input data to output file. Use original data 2) Internally refmac groups reflection into classes. All twin related reflections belong to one class. In the example you give four reflections belong to one class. FreeR is property of the class not individual relfections. I.e. all twin related reflections belong either to free or working class. If your input data has this property then output should also have this. In the output file there will be 0 (for free) and 1 (working) 3) At early stages it is not easy to factorise twin and underestimated symmetry. It seems that L-test is the best way of making decision if you crystals are twinnined. If you collect data from many crystals and merge them then you artificially twin (or increase twinning). Using pointless to index all datasets consistently may sort some of the problems 4) In this case it seems that you may need to reindex. Largest twin domain is not the first one Regards Garib On 2 Mar 2012, at 02:00, wtempel wrote: Dear CCp4ers, A good morning to everyone. Today, I have a structure that I initially refined in space group P6522, 1mol/asu. Scaling stats (scalepack): 2.30-2.26A: Rsym=99.9%; I/sigma 3 2.61-2.55A: Rsym=39.6%, I/sigma 10 50.00-6.13: Rsym=6.4% Some mild anisotropy in the resolution limits is apparent on the diffraction images. Say, visible spots at 2.2A in one direction, 2.6A in the other. Rfree, using data to 2.3A, was stuck in mid-30%s. The map appears like 3.5A resolution, with some difference density for loops that cannot be interpreted with reasonable geometry. Rsym is very similar for data scaled in P3, in all resolution shells. Xtriage does not suggest merohedral twinning. Nevertheless, I extended my free flags in sftools from P6522 to P32 and cad'd them to amplitudes merged in spacegroup P32. Correspondingly, I expanded my model to a homotetramer and ran Refmac with amplitude based twinning. (Would this be a reasonable input to twin refinement?) From the output coordinates: REMARK 3 TWIN DETAILS REMARK 3 NUMBER OF TWIN DOMAINS :4 REMARK 3 TWIN DOMAIN :1 REMARK 3 TWIN OPERATOR : H, K, L REMARK 3 TWIN FRACTION : 0.269 REMARK 3 TWIN DOMAIN :2 REMARK 3 TWIN OPERATOR : -K, -H, -L REMARK 3 TWIN FRACTION : 0.171 REMARK 3 TWIN DOMAIN :3 REMARK 3 TWIN OPERATOR : K, H, -L REMARK 3 TWIN FRACTION : 0.258 REMARK 3 TWIN DOMAIN :4 REMARK 3 TWIN OPERATOR : -H, -K, L REMARK 3 TWIN FRACTION : 0.302 Does this establish twinning versus underestimated symmetry? And what do I need to know about my free-R? Did refmac assign a new flag? Whereas the output file's flags are all 1s and 0s, the input file had 0 ... 19. During the first run, Rfree dropped to 28%. But on a subsequent run, Rfree was stuck 30% when I used the initial job's output MTZ. Many thanks in advance for your helpful comments. Wolfram Tempel Garib N Murshudov Structural Studies Division MRC Laboratory of Molecular Biology Hills Road Cambridge CB2 0QH UK Email: ga...@mrc-lmb.cam.ac.uk Web http://www.mrc-lmb.cam.ac.uk
Re: [ccp4bb] twin refinement in refmac
On 2 Mar 2012, at 08:01, arka chakraborty wrote: Hi all, I will like to know, as a follow up of what Prof. Randy Read said, what should be done to do the refinement against the measured data and not the detwinned F( which refmac outputs in the mtz after twin refinement), during subsequent refinements. Operationally, don't put in the MTZ in field of the GUI something that Refmac generated for you in a previous run as MTZ out file. Always use as MTZ in your original data file. And also, I would like to know how to ensure that the free R generated takes twinning into account if I am not using phenix. Refmac does take in consideration the twin law(s) when handling free reflections . This is the case even if you have generated your free reflections randomly. Internally Refmac will modify your Free set in such a way that twin related reflections are in the same group (free or working) -- classes mentioned by Garib The good thing about this is that twin-related reflections are handled properly during refinement irrespective of your Free-set choice. The bad thing (Garib please correct me if I am wrong here) is that you might end up depositing a Free set which is not that actually used in refinement. Best wishes Roberto Thanks in advance, Regards, ARKO On Fri, Mar 2, 2012 at 12:44 PM, Randy J. Read rj...@cam.ac.ukmailto:rj...@cam.ac.uk wrote: I'm worried when you say that you use the initial job's output MTZ. Refmac replaces F with a detwinned F in the output file so you wouldn't be refining against your measured data in the subsequent round. Best wishes Randy Read Randy J. Read On 2 Mar 2012, at 02:00, wtempel wtem...@gmail.commailto:wtem...@gmail.com wrote: Dear CCp4ers, A good morning to everyone. Today, I have a structure that I initially refined in space group P6522, 1mol/asu. Scaling stats (scalepack): 2.30-2.26A: Rsym=99.9%; I/sigma 3 2.61-2.55A: Rsym=39.6%, I/sigma 10 50.00-6.13: Rsym=6.4% Some mild anisotropy in the resolution limits is apparent on the diffraction images. Say, visible spots at 2.2A in one direction, 2.6A in the other. Rfree, using data to 2.3A, was stuck in mid-30%s. The map appears like 3.5A resolution, with some difference density for loops that cannot be interpreted with reasonable geometry. Rsym is very similar for data scaled in P3, in all resolution shells. Xtriage does not suggest merohedral twinning. Nevertheless, I extended my free flags in sftools from P6522 to P32 and cad'd them to amplitudes merged in spacegroup P32. Correspondingly, I expanded my model to a homotetramer and ran Refmac with amplitude based twinning. (Would this be a reasonable input to twin refinement?) From the output coordinates: REMARK 3 TWIN DETAILS REMARK 3 NUMBER OF TWIN DOMAINS :4 REMARK 3 TWIN DOMAIN :1 REMARK 3 TWIN OPERATOR : H, K, L REMARK 3 TWIN FRACTION : 0.269 REMARK 3 TWIN DOMAIN :2 REMARK 3 TWIN OPERATOR : -K, -H, -L REMARK 3 TWIN FRACTION : 0.171 REMARK 3 TWIN DOMAIN :3 REMARK 3 TWIN OPERATOR : K, H, -L REMARK 3 TWIN FRACTION : 0.258 REMARK 3 TWIN DOMAIN :4 REMARK 3 TWIN OPERATOR : -H, -K, L REMARK 3 TWIN FRACTION : 0.302 Does this establish twinning versus underestimated symmetry? And what do I need to know about my free-R? Did refmac assign a new flag? Whereas the output file's flags are all 1s and 0s, the input file had 0 ... 19. During the first run, Rfree dropped to 28%. But on a subsequent run, Rfree was stuck 30% when I used the initial job's output MTZ. Many thanks in advance for your helpful comments. Wolfram Tempel -- ARKA CHAKRABORTY CAS in Crystallography and Biophysics University of Madras Chennai,India Roberto Steiner, PhD Group Leader Randall Division of Cell and Molecular Biophysics King's College London Room 3.10A New Hunt's House Guy's Campus SE1 1UL, London, UK Tel 0044-20-78488216 Fax 0044-20-78486435 roberto.stei...@kcl.ac.ukmailto:roberto.stei...@kcl.ac.uk
Re: [ccp4bb] twin refinement in refmac
Hi all, Thanks for the express replies. Your insights along with the article by Prof. Garib pointed to by Prof. Pavel completes the story for me. Regards, ARKO On Fri, Mar 2, 2012 at 3:09 PM, Steiner, Roberto roberto.stei...@kcl.ac.ukwrote: On 2 Mar 2012, at 08:01, arka chakraborty wrote: Hi all, I will like to know, as a follow up of what Prof. Randy Read said, what should be done to do the refinement against the measured data and not the detwinned F( which refmac outputs in the mtz after twin refinement), during subsequent refinements. Operationally, don't put in the MTZ in field of the GUI something that Refmac generated for you in a previous run as MTZ out file. Always use as MTZ in your original data file. And also, I would like to know how to ensure that the free R generated takes twinning into account if I am not using phenix. Refmac does take in consideration the twin law(s) when handling free reflections . This is the case even if you have generated your free reflections randomly. Internally Refmac will modify your Free set in such a way that twin related reflections are in the same group (free or working) -- classes mentioned by Garib The good thing about this is that twin-related reflections are handled properly during refinement irrespective of your Free-set choice. The bad thing (Garib please correct me if I am wrong here) is that you might end up depositing a Free set which is not that actually used in refinement. Best wishes Roberto Thanks in advance, Regards, ARKO On Fri, Mar 2, 2012 at 12:44 PM, Randy J. Read rj...@cam.ac.uk wrote: I'm worried when you say that you use the initial job's output MTZ. Refmac replaces F with a detwinned F in the output file so you wouldn't be refining against your measured data in the subsequent round. Best wishes Randy Read Randy J. Read On 2 Mar 2012, at 02:00, wtempel wtem...@gmail.com wrote: Dear CCp4ers, A good morning to everyone. Today, I have a structure that I initially refined in space group P6522, 1mol/asu. Scaling stats (scalepack): 2.30-2.26A: Rsym=99.9%; I/sigma 3 2.61-2.55A: Rsym=39.6%, I/sigma 10 50.00-6.13: Rsym=6.4% Some mild anisotropy in the resolution limits is apparent on the diffraction images. Say, visible spots at 2.2A in one direction, 2.6A in the other. Rfree, using data to 2.3A, was stuck in mid-30%s. The map appears like 3.5A resolution, with some difference density for loops that cannot be interpreted with reasonable geometry. Rsym is very similar for data scaled in P3, in all resolution shells. Xtriage does not suggest merohedral twinning. Nevertheless, I extended my free flags in sftools from P6522 to P32 and cad'd them to amplitudes merged in spacegroup P32. Correspondingly, I expanded my model to a homotetramer and ran Refmac with amplitude based twinning. (Would this be a reasonable input to twin refinement?) From the output coordinates: REMARK 3 TWIN DETAILS REMARK 3 NUMBER OF TWIN DOMAINS :4 REMARK 3 TWIN DOMAIN :1 REMARK 3 TWIN OPERATOR : H, K, L REMARK 3 TWIN FRACTION : 0.269 REMARK 3 TWIN DOMAIN :2 REMARK 3 TWIN OPERATOR : -K, -H, -L REMARK 3 TWIN FRACTION : 0.171 REMARK 3 TWIN DOMAIN :3 REMARK 3 TWIN OPERATOR : K, H, -L REMARK 3 TWIN FRACTION : 0.258 REMARK 3 TWIN DOMAIN :4 REMARK 3 TWIN OPERATOR : -H, -K, L REMARK 3 TWIN FRACTION : 0.302 Does this establish twinning versus underestimated symmetry? And what do I need to know about my free-R? Did refmac assign a new flag? Whereas the output file's flags are all 1s and 0s, the input file had 0 ... 19. During the first run, Rfree dropped to 28%. But on a subsequent run, Rfree was stuck 30% when I used the initial job's output MTZ. Many thanks in advance for your helpful comments. Wolfram Tempel -- *ARKA CHAKRABORTY* *CAS in Crystallography and Biophysics* *University of Madras* *Chennai,India* Roberto Steiner, PhD Group Leader Randall Division of Cell and Molecular Biophysics King's College London Room 3.10A New Hunt's House Guy's Campus SE1 1UL, London, UK Tel 0044-20-78488216 Fax 0044-20-78486435 roberto.stei...@kcl.ac.uk -- *ARKA CHAKRABORTY* *CAS in Crystallography and Biophysics* *University of Madras* *Chennai,India*
[ccp4bb] twin refinement in refmac
Dear CCp4ers, A good morning to everyone. Today, I have a structure that I initially refined in space group P6522, 1mol/asu. Scaling stats (scalepack): 2.30-2.26A: Rsym=99.9%; I/sigma 3 2.61-2.55A: Rsym=39.6%, I/sigma 10 50.00-6.13: Rsym=6.4% Some mild anisotropy in the resolution limits is apparent on the diffraction images. Say, visible spots at 2.2A in one direction, 2.6A in the other. Rfree, using data to 2.3A, was stuck in mid-30%s. The map appears like 3.5A resolution, with some difference density for loops that cannot be interpreted with reasonable geometry. Rsym is very similar for data scaled in P3, in all resolution shells. Xtriage does not suggest merohedral twinning. Nevertheless, I extended my free flags in sftools from P6522 to P32 and cad'd them to amplitudes merged in spacegroup P32. Correspondingly, I expanded my model to a homotetramer and ran Refmac with amplitude based twinning. (Would this be a reasonable input to twin refinement?) From the output coordinates: REMARK 3 TWIN DETAILS REMARK 3 NUMBER OF TWIN DOMAINS :4 REMARK 3 TWIN DOMAIN :1 REMARK 3 TWIN OPERATOR : H, K, L REMARK 3 TWIN FRACTION : 0.269 REMARK 3 TWIN DOMAIN :2 REMARK 3 TWIN OPERATOR : -K, -H, -L REMARK 3 TWIN FRACTION : 0.171 REMARK 3 TWIN DOMAIN :3 REMARK 3 TWIN OPERATOR : K, H, -L REMARK 3 TWIN FRACTION : 0.258 REMARK 3 TWIN DOMAIN :4 REMARK 3 TWIN OPERATOR : -H, -K, L REMARK 3 TWIN FRACTION : 0.302 Does this establish twinning versus underestimated symmetry? And what do I need to know about my free-R? Did refmac assign a new flag? Whereas the output file's flags are all 1s and 0s, the input file had 0 ... 19. During the first run, Rfree dropped to 28%. But on a subsequent run, Rfree was stuck 30% when I used the initial job's output MTZ. Many thanks in advance for your helpful comments. Wolfram Tempel
Re: [ccp4bb] twin refinement in refmac
Hi Wolfram, a few points: - R-factors in twin refinement vs non-twin refinement are not directly comparable: G.N. Murshudov, Appl. Comput. Math., V.10, N.2, 2011, pp.250-261 http://www.science.az/acm/V10,%20N2,%202011,%20pdf/250-261.pdf - did you make sure free-R flags assigned having twinning in mind (what phenix.refine always does)? Otherwise you have a risk of having this: see page 14 here: http://phenix-online.org/presentations/latest/pavel_validation.pdf Pavel On Thu, Mar 1, 2012 at 6:00 PM, wtempel wtem...@gmail.com wrote: Dear CCp4ers, A good morning to everyone. Today, I have a structure that I initially refined in space group P6522, 1mol/asu. Scaling stats (scalepack): 2.30-2.26A: Rsym=99.9%; I/sigma 3 2.61-2.55A: Rsym=39.6%, I/sigma 10 50.00-6.13: Rsym=6.4% Some mild anisotropy in the resolution limits is apparent on the diffraction images. Say, visible spots at 2.2A in one direction, 2.6A in the other. Rfree, using data to 2.3A, was stuck in mid-30%s. The map appears like 3.5A resolution, with some difference density for loops that cannot be interpreted with reasonable geometry. Rsym is very similar for data scaled in P3, in all resolution shells. Xtriage does not suggest merohedral twinning. Nevertheless, I extended my free flags in sftools from P6522 to P32 and cad'd them to amplitudes merged in spacegroup P32. Correspondingly, I expanded my model to a homotetramer and ran Refmac with amplitude based twinning. (Would this be a reasonable input to twin refinement?) From the output coordinates: REMARK 3 TWIN DETAILS REMARK 3 NUMBER OF TWIN DOMAINS :4 REMARK 3 TWIN DOMAIN :1 REMARK 3 TWIN OPERATOR : H, K, L REMARK 3 TWIN FRACTION : 0.269 REMARK 3 TWIN DOMAIN :2 REMARK 3 TWIN OPERATOR : -K, -H, -L REMARK 3 TWIN FRACTION : 0.171 REMARK 3 TWIN DOMAIN :3 REMARK 3 TWIN OPERATOR : K, H, -L REMARK 3 TWIN FRACTION : 0.258 REMARK 3 TWIN DOMAIN :4 REMARK 3 TWIN OPERATOR : -H, -K, L REMARK 3 TWIN FRACTION : 0.302 Does this establish twinning versus underestimated symmetry? And what do I need to know about my free-R? Did refmac assign a new flag? Whereas the output file's flags are all 1s and 0s, the input file had 0 ... 19. During the first run, Rfree dropped to 28%. But on a subsequent run, Rfree was stuck 30% when I used the initial job's output MTZ. Many thanks in advance for your helpful comments. Wolfram Tempel
Re: [ccp4bb] twin refinement in refmac
I'm worried when you say that you use the initial job's output MTZ. Refmac replaces F with a detwinned F in the output file so you wouldn't be refining against your measured data in the subsequent round. Best wishes Randy Read Randy J. Read On 2 Mar 2012, at 02:00, wtempel wtem...@gmail.com wrote: Dear CCp4ers, A good morning to everyone. Today, I have a structure that I initially refined in space group P6522, 1mol/asu. Scaling stats (scalepack): 2.30-2.26A: Rsym=99.9%; I/sigma 3 2.61-2.55A: Rsym=39.6%, I/sigma 10 50.00-6.13: Rsym=6.4% Some mild anisotropy in the resolution limits is apparent on the diffraction images. Say, visible spots at 2.2A in one direction, 2.6A in the other. Rfree, using data to 2.3A, was stuck in mid-30%s. The map appears like 3.5A resolution, with some difference density for loops that cannot be interpreted with reasonable geometry. Rsym is very similar for data scaled in P3, in all resolution shells. Xtriage does not suggest merohedral twinning. Nevertheless, I extended my free flags in sftools from P6522 to P32 and cad'd them to amplitudes merged in spacegroup P32. Correspondingly, I expanded my model to a homotetramer and ran Refmac with amplitude based twinning. (Would this be a reasonable input to twin refinement?) From the output coordinates: REMARK 3 TWIN DETAILS REMARK 3 NUMBER OF TWIN DOMAINS :4 REMARK 3 TWIN DOMAIN :1 REMARK 3 TWIN OPERATOR : H, K, L REMARK 3 TWIN FRACTION : 0.269 REMARK 3 TWIN DOMAIN :2 REMARK 3 TWIN OPERATOR : -K, -H, -L REMARK 3 TWIN FRACTION : 0.171 REMARK 3 TWIN DOMAIN :3 REMARK 3 TWIN OPERATOR : K, H, -L REMARK 3 TWIN FRACTION : 0.258 REMARK 3 TWIN DOMAIN :4 REMARK 3 TWIN OPERATOR : -H, -K, L REMARK 3 TWIN FRACTION : 0.302 Does this establish twinning versus underestimated symmetry? And what do I need to know about my free-R? Did refmac assign a new flag? Whereas the output file's flags are all 1s and 0s, the input file had 0 ... 19. During the first run, Rfree dropped to 28%. But on a subsequent run, Rfree was stuck 30% when I used the initial job's output MTZ. Many thanks in advance for your helpful comments. Wolfram Tempel
Re: [ccp4bb] twin refinement
Hi, You don't say whether you've considered the possibility that the true symmetry is higher than P61, e.g. P6122. If there's higher symmetry consistent with your data, then either pointless or xtriage will tell you which space groups to consider for test refinements. Another good test (if you haven't allowed NCS-related copies to diverge in the refinement in P61) is the R vs R test, to tell whether the apparent twin operator is really a symmetry operator in the model. Note that the R-factors aren't the best indicator, as they tend to drop when you assume perfect twinning, even if it doesn't exist. I think Garib has some of his talks on the web, giving some indication of what you can expect. Regards, Randy Read On 3 Mar 2010, at 04:35, subhash C bihani wrote: Hi i am refining a protein strucutre with space group P61. Intesity statistics doesnot suggest any twining but refmac and phenix xtriage suggest a twin operator with fraction 0.48. when refined with this twin operator R factors come dowm by 3-5% and also map looks much better. but upon closer inspceion it seems that maps are heavily biased toward the model in both refmac and phenix ( I have mutated one lysine to his and it changed the map to cover his). my question is why twin refinement maps are biased and what are the ways to overcome this. my second question is how to decide whether to use twin refinement or not. is it depends on R factor statistics or we have to check some other parameters. thanks in advance subhash bihani scientific officer Bhabha Atomic Research Centre Trombay-400085 Mumbai(M.S.) India Ph. +912594688 (o) +919322191020 subhash bihani scientific officer Bhabha Atomic Research Centre Trombay-400085 Mumbai(M.S.) India Ph. +912594688 (o) +919322191020 subhash bihani scientific officer Bhabha Atomic Research Centre Trombay-400085 Mumbai(M.S.) India Ph. +912594688 (o) +919322191020 - -- Randy J. Read Department of Haematology, University of Cambridge Cambridge Institute for Medical Research Tel: + 44 1223 336500 Wellcome Trust/MRC Building Fax: + 44 1223 336827 Hills RoadE-mail: rj...@cam.ac.uk Cambridge CB2 0XY, U.K. www-structmed.cimr.cam.ac.uk
[ccp4bb] twin refinement
Hi i am refining a protein strucutre with space group P61. Intesity statistics doesnot suggest any twining but refmac and phenix xtriage suggest a twin operator with fraction 0.48. when refined with this twin operator R factors come dowm by 3-5% and also map looks much better. but upon closer inspceion it seems that maps are heavily biased toward the model in both refmac and phenix ( I have mutated one lysine to his and it changed the map to cover his). my question is why twin refinement maps are biased and what are the ways to overcome this. my second question is how to decide whether to use twin refinement or not. is it depends on R factor statistics or we have to check some other parameters. thanks in advance subhash bihani scientific officer Bhabha Atomic Research Centre Trombay-400085 Mumbai(M.S.) India Ph. +912594688 (o) +919322191020 subhash bihani scientific officer Bhabha Atomic Research Centre Trombay-400085 Mumbai(M.S.) India Ph. +912594688 (o) +919322191020 subhash bihani scientific officer Bhabha Atomic Research Centre Trombay-400085 Mumbai(M.S.) India Ph. +912594688 (o) +919322191020 -
Re: [ccp4bb] Twin refinement not working in CCP4 6.1.1
Dear All this and other fixes are on the ccp4 problems pages http://www.ccp4.ac.uk/problems.php Charles Ballard On 2 Mar 2009, at 16:50, Ronnie Berntsson wrote: I have the same problem when running the latest ccp4 (6.1.1 on os x, updated 1 hour ago via fink). Only way I managed to circumvent it is to add the TWIN keyword in the command file as well.. Cheers, Ronnie Berntsson On Mar 2, 2009, at 16:06, Carr, SB (Stephen) wrote: Dear CCP4BB, I have come across a problem when running refmac5 in CCP4 6.1.1 via the gui. I have a twinned data set and so want to refine the twin fraction of the data. I selected amplitude based twin refinement in the gui, but find that the resulting command file does not contain the twin keyword and hence twin refinement does not occur. If I manually edit the command file to include the twin keyword then refmac calculates twin fractions etc. Has anyone else come across this problem and if so how did you solve it? best regards, Steve Carr Dr Stephen Carr, Membrane Protein Laboratory, Diamond Light Source Ltd, Harwell Science and Innovation Campus, Chilton, Didcot, OX11 0DE Tel:(44)1235 778896 Email stephen.c...@diamond.ac.uk This e-mail and any attachments may contain confidential, copyright and or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail. Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd. Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be transmitted in or with the message. Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom Scanned by iCritical.
Re: [ccp4bb] Twin refinement not working in CCP4 6.1.1
I have the same problem when running the latest ccp4 (6.1.1 on os x, updated 1 hour ago via fink). Only way I managed to circumvent it is to add the TWIN keyword in the command file as well.. Cheers, Ronnie Berntsson On Mar 2, 2009, at 16:06, Carr, SB (Stephen) wrote: Dear CCP4BB, I have come across a problem when running refmac5 in CCP4 6.1.1 via the gui. I have a twinned data set and so want to refine the twin fraction of the data. I selected amplitude based twin refinement in the gui, but find that the resulting command file does not contain the twin keyword and hence twin refinement does not occur. If I manually edit the command file to include the twin keyword then refmac calculates twin fractions etc. Has anyone else come across this problem and if so how did you solve it? best regards, Steve Carr Dr Stephen Carr, Membrane Protein Laboratory, Diamond Light Source Ltd, Harwell Science and Innovation Campus, Chilton, Didcot, OX11 0DE Tel:(44)1235 778896 Email stephen.c...@diamond.ac.uk This e-mail and any attachments may contain confidential, copyright and or privileged material, and are for the use of the intended addressee only. If you are not the intended addressee or an authorised recipient of the addressee please notify us of receipt by returning the e-mail and do not use, copy, retain, distribute or disclose the information in or attached to the e-mail. Any opinions expressed within this e-mail are those of the individual and not necessarily of Diamond Light Source Ltd. Diamond Light Source Ltd. cannot guarantee that this e-mail or any attachments are free from viruses and we cannot accept liability for any damage which you may sustain as a result of software viruses which may be transmitted in or with the message. Diamond Light Source Limited (company no. 4375679). Registered in England and Wales with its registered office at Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire, OX11 0DE, United Kingdom Scanned by iCritical.
Re: [ccp4bb] Twin refinement - selecting FreeR Q for Garib..
Is it true that when doing twinned refinement REFMAC does not use the assigned FreeR set? Eleanor Roberto Steiner wrote: Hi Sabine, If your question is: Is it possible to refine twinned structures with Refmac? My answer is: Yes. Very well. Best wishes, Roberto On 20 Feb 2009, at 16:32, Sabine Schneider wrote: Hi everyone, I got good data to 2.4A on a protein-ligand complex. I want to solve the structure by MR using a model with 45% seq. ID and 55% similarity. Initially the data appeared to be P622 (pointless, selfrotation function etc). Unit cell: 145 145 65, MW protein is 28kDa I prepared the MR model with chainsaw, choping of non-conserved residues at the C gamma and reset the B-factors. Phaser found a solution, but with negativ LLG. I than processed the data in P3, P321, P312 and P6 and run Phaser searching all possible alternative spacegroups. The best solution is P3 (4mol/asu) followed by P321 (2mol/asu),both with a LLG of above 400. But when trying to refine the models in Refmac the R/Rfree stays at ~48%. Looking at the truncate output and phenix xtriage, twinning is suggested with the merohedral twin law -h, -k, l and a twin fraction of 40.3% Using twin refine option in Refmac (5.5.0070), the R/Rfree for the P321 solution drops to around 33%, but the difference between R/Rfree is only 1%. For the solution found in P3 the R/Rfree stays at around 47%. So I assumed that P321 is the better solution. Is the difference in R/Rfree only 1%, because the free and work reflections are related through the twin law? I used phenix to assign the free reflections putting in the twin operators. Doing simulated annealing in phenix, I get a rather large difference in R/Rfree of ~7%. Well, I guess I need to do some tweaking of the parameters in phenix (running the latest phenix and cci_apps). When I than use these free reflections assigned by phenix in Refmac I still get only 1.5% difference between R and Rfree? So is it doable in Refmac, or is my best bet phenix? Any advice what is the best way to proceed is much appreciated! Sabine -- Dr. Sabine Schneider Ludwig-Maximilians-University Department of Chemistry and Pharmacy Butenandtstrasse 5-13, Building F 81377 Munich Germany Phone: +49 (0)89 2180 77846 Fax: +49 (0)89 2180 77756 http://www.carellgroup.de/ --- Roberto Steiner Randall Division of Cell and Molecular Biophysics New Hunt's House King's College London Guy's Campus London, SE1 1UL Phone +44 (0)20-7848-8216 Fax +44 (0)20-7848-6435 e-mail roberto.stei...@kcl.ac.uk
[ccp4bb] Twin refinement - selecting FreeR
Hi everyone, I got good data to 2.4A on a protein-ligand complex. I want to solve the structure by MR using a model with 45% seq. ID and 55% similarity. Initially the data appeared to be P622 (pointless, selfrotation function etc). Unit cell: 145 145 65, MW protein is 28kDa I prepared the MR model with chainsaw, choping of non-conserved residues at the C gamma and reset the B-factors. Phaser found a solution, but with negativ LLG. I than processed the data in P3, P321, P312 and P6 and run Phaser searching all possible alternative spacegroups. The best solution is P3 (4mol/asu) followed by P321 (2mol/asu),both with a LLG of above 400. But when trying to refine the models in Refmac the R/Rfree stays at ~48%. Looking at the truncate output and phenix xtriage, twinning is suggested with the merohedral twin law -h, -k, l and a twin fraction of 40.3% Using twin refine option in Refmac (5.5.0070), the R/Rfree for the P321 solution drops to around 33%, but the difference between R/Rfree is only 1%. For the solution found in P3 the R/Rfree stays at around 47%. So I assumed that P321 is the better solution. Is the difference in R/Rfree only 1%, because the free and work reflections are related through the twin law? I used phenix to assign the free reflections putting in the twin operators. Doing simulated annealing in phenix, I get a rather large difference in R/Rfree of ~7%. Well, I guess I need to do some tweaking of the parameters in phenix (running the latest phenix and cci_apps). When I than use these free reflections assigned by phenix in Refmac I still get only 1.5% difference between R and Rfree? So is it doable in Refmac, or is my best bet phenix? Any advice what is the best way to proceed is much appreciated! Sabine -- Dr. Sabine Schneider Ludwig-Maximilians-University Department of Chemistry and Pharmacy Butenandtstrasse 5-13, Building F 81377 Munich Germany Phone: +49 (0)89 2180 77846 Fax: +49 (0)89 2180 77756 http://www.carellgroup.de/
Re: [ccp4bb] Twin refinement - selecting FreeR
Hi Sabine, If your question is: Is it possible to refine twinned structures with Refmac? My answer is: Yes. Very well. Best wishes, Roberto On 20 Feb 2009, at 16:32, Sabine Schneider wrote: Hi everyone, I got good data to 2.4A on a protein-ligand complex. I want to solve the structure by MR using a model with 45% seq. ID and 55% similarity. Initially the data appeared to be P622 (pointless, selfrotation function etc). Unit cell: 145 145 65, MW protein is 28kDa I prepared the MR model with chainsaw, choping of non-conserved residues at the C gamma and reset the B-factors. Phaser found a solution, but with negativ LLG. I than processed the data in P3, P321, P312 and P6 and run Phaser searching all possible alternative spacegroups. The best solution is P3 (4mol/asu) followed by P321 (2mol/asu),both with a LLG of above 400. But when trying to refine the models in Refmac the R/Rfree stays at ~48%. Looking at the truncate output and phenix xtriage, twinning is suggested with the merohedral twin law -h, -k, l and a twin fraction of 40.3% Using twin refine option in Refmac (5.5.0070), the R/Rfree for the P321 solution drops to around 33%, but the difference between R/Rfree is only 1%. For the solution found in P3 the R/Rfree stays at around 47%. So I assumed that P321 is the better solution. Is the difference in R/Rfree only 1%, because the free and work reflections are related through the twin law? I used phenix to assign the free reflections putting in the twin operators. Doing simulated annealing in phenix, I get a rather large difference in R/Rfree of ~7%. Well, I guess I need to do some tweaking of the parameters in phenix (running the latest phenix and cci_apps). When I than use these free reflections assigned by phenix in Refmac I still get only 1.5% difference between R and Rfree? So is it doable in Refmac, or is my best bet phenix? Any advice what is the best way to proceed is much appreciated! Sabine -- Dr. Sabine Schneider Ludwig-Maximilians-University Department of Chemistry and Pharmacy Butenandtstrasse 5-13, Building F 81377 Munich Germany Phone: +49 (0)89 2180 77846 Fax: +49 (0)89 2180 77756 http://www.carellgroup.de/ --- Roberto Steiner Randall Division of Cell and Molecular Biophysics New Hunt's House King's College London Guy's Campus London, SE1 1UL Phone +44 (0)20-7848-8216 Fax +44 (0)20-7848-6435 e-mail roberto.stei...@kcl.ac.uk
Re: [ccp4bb] Twin refinement - selecting FreeR
I used phenix to assign the free reflections putting in the twin operators. Doing simulated annealing in phenix, I get a rather large difference in R/Rfree of ~7%. Well, I guess I need to do some tweaking of the parameters in phenix (running the latest phenix and cci_apps). When I than use these free reflections assigned by phenix in Refmac I still get only 1.5% difference between R and Rfree? Are you certain that you're using the same test set for both programs? Refmac expects that 0 marks the test set. Phenix will recognize this and adjust to use it automatically, but by default it still uses the Xplor/CNS convention where the test set is marked 1 and everything else is 0. If you flagged reflections with Phenix, it's possible that Refmac is using your test reflections as your working reflections. (I suspect that you'd see other problems too if this was the case, but I've made this same mistake before and I think it resulted in very close Rwork/Rfree.) -Nat
[ccp4bb] Twin refinement with SHELXL
Dear All, I have 1.4 A data and a molecular replacement solution for a crystal indexed as C2, with beta approximately equal to 90. Refinement with refmac is progressing poorly, and intensity statistics (Truncate) and other twinning tests (xtriage) suggest pseudo-merohedral twinning with a twin fraction around 0.44. I want to try refining in SHELXL with a twin-specific target function, but I have some questions about how to get started. SHELX documentation indicates that the reflection file should contain intensities rather than amplitudes, and preferably unmerged data. I have been refining against an mtz file that was integrated with MOSFLM, scaled and merged with SCALA, converted to amplitudes with TRUNCATE, then had freeR flags assigned. I am getting the mtz2various patch recently mentioned on the bb, that will make an mtz file into a proper SHELX hkl file, but which mtz file do I convert? The mtz output by TRUNCATE contains both F's and I's, so I could use that one to get measured intensities, or do I actually need to go back further to get my unmerged reflections? A second question relates to Rfree flags - should I transfer the ones I've been using up until now? They were a randomly assigned 5%. Now I've read that using the thin shell method is better for twinned crystals and NCS, but I don't know exactly how that is done... any recommendations? 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)
