- phenixbb - phenix-online.org

Fwd: Re: anisotropic induced noise amplification
by Pavel Afonine 11 May '12

11 May '12

Hi, I realized Randy's reply did not go to the mailing list, so I'm forwarding it now. Pavel -------- Original Message -------- Subject: Re: [phenixbb] anisotropic induced noise amplification Date: Mon, 7 May 2012 09:49:57 +0100 From: Randy Read <rjr27(a)cam.ac.uk> To: Pavel Afonine <pafonine(a)lbl.gov> CC: developers(a)phenix-online.org <developers(a)phenix-online.org>, Peter Zwart <PHZwart(a)lbl.gov> In the absence of an answer from Peter (who implemented this and therefore ought to know best), I'm pretty sure that the idea is to measure how much difference there is in the amount of signal in the data that have received the biggest anisotropy corrections, compared to the ones that have received the smallest anisotropy corrections. If there's very little signal in the data that have been scaled up the most, then the noise is being amplified (nearly-random amplitudes have gone from being relatively small to the same size on average as the well-measured amplitudes). But what the Z-scores mean is not at all clear from that output, and I'm not aware that Peter has written anything describing this. This amplified noise is what I suspect is creating problems in phenix.refine refinement with highly anisotropic data, because if the experimental sigmas are ignored then the program doesn't know (after anisotropic scaling) which amplitudes of a comparable size contain signal and which are essentially noise. Anisotropic truncation gets rid of it, which would explain why refinement works better with anisotropically-truncated data. Randy On 6 May 2012, at 20:21, Pavel Afonine wrote: > Hi, > > did anyone address this post? I'm interested to know the answer > myself: I have absolutely no idea what is this. > > Pavel > > -------- Original Message -------- > Subject: [phenixbb] anisotropic induced noise amplification > Date: Wed, 2 May 2012 16:06:55 -0500 > From: James Thompson <Thompson.James(a)mayo.edu> > Reply-To: PHENIX user mailing list <phenixbb(a)phenix-online.org> > To: phenixbb(a)phenix-online.org > > > > Anyone willing to summarize Xtriage's algorithm that determines > whether "/anisotropic induced noise amplification/" is present in > diffraction data? What is this noise amplification? Is it > anisotropic correction factors that over-fit the data thereby inducing > noise? > > Many thanks, > Jim T > > ---------------- Anisotropy analyses ---------------- > > Anisotropy ( [MaxAnisoB-MinAnisoB]/[MaxAnisoB] ) : 3.596e-01 > Anisotropic ratio p-value : 0.000e+00 > > The p-value is a measure of the severity of anisotropy as > observed in the PDB. > The p-value of 0.000e+00 indicates that roughly 100.0 % of > datasets available in the PDB have > an anisotropy equal to or worse than this dataset. > > > For the resolution shell spanning between 2.36 - 2.20 Angstrom, > the mean I/sigI is equal to 3.37. 42.2 % of these intensities have > an I/sigI > 3. When sorting these intensities by their anisotropic > correction factor and analysing the I/sigI behavior for this ordered > list, we can gauge the presence of 'anisotropy induced noise > amplification' > in reciprocal space. > > The quarter of Intensities *least* affected by the anisotropy > correction show > <I/sigI> : 3.82e+00 > Fraction of I/sigI > 3 : 5.01e-01 ( Z = 3.59 ) > > The quarter of Intensities *most* affected by the anisotropy > correction show > <I/sigI> : 2.02e+00 > Fraction of I/sigI > 3 : 1.84e-01 ( Z = 10.78 ) > > *The combined Z-score of 11.36 indicates that there probably is > significant* > *systematic noise amplification that could possibly lead to artefacts > in the* > *maps or difficulties in refinement* > > <Attached Message Part.txt> ------ 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 Road E-mail: rjr27(a)cam.ac.uk <mailto:[email protected]> Cambridge CB2 0XY, U.K. www-structmed.cimr.cam.ac.uk

1 0

lastest phenix.refine reproducible problem
by 周文昌 09 May '12

09 May '12

Dear all, I've been learning phenix refinement on a MAD dataset which is already deposited in PDB. There are something I can not get through. Basically I had two steps refinement: 1->phenix.refine data.mtz model.pdb xray_data.high_resolution=3 fix_rotamers=true strategy=individual+adp+individual_sites+tls 2->phenix.refine data.mtz model-refined.pdb xray_data.high_resolution=3 fix_rotamers=false strategy=none main.number_of_macro_cycles=1 model-refined.pdb is actually from the 1st step, so in principle the R/R-free values should be the same for 1 and 2. However in my case, I had 0.2506/0.2996 for the 1st step but 0.2503/0.2998 for the second. I compared the resulting pdb and mtz files from the two steps. PDB model were exactly the same but there were little difference in the mtz file. Why there is difference in mtz file and would it be the reason for the difference of R factors? I use the latest version of phenix, phenix-dev-1009. thanks, Wenchang

3 3

Installation on a POWER 7 system
by Rajagopalan, Senapathy 09 May '12

09 May '12

Hi Everyone, I was wondering if its possible to install PHENIX on a POWER 7 system. The cluster for which we have access runs on a power 7 system and at least my initial try to install the FC3 version failed. 'uname –rm' gives me '2.6.32-220.13.1.el6.ppc64 ppc64'. If only re-compilation is required, I would appreciate if you can point me in the right direction. Thanks Sena Methodist. Leading Medicine. Recognized by U.S.News & World Report as one of America's "Best Hospitals" in 13 specialties. Named to FORTUNE® Magazine's "100 Best Companies to Work For®" list seven years in a row. Designated as a Magnet hospital for excellence in nursing. Visit us at methodisthealth.com. Follow us at twitter.com/MethodistHosp and www.facebook.com/MethodistHospital. ***CONFIDENTIALITY NOTICE*** This e-mail is the property of The Methodist Hospital and/or its relevant affiliates and may contain restricted and privileged material for the sole use of the intended recipient(s). Any review, use, distribution or disclosure by others is strictly prohibited. If you are not the intended recipient (or authorized to receive for the recipient), please contact the sender and delete all copies of the message. Thank you.

2 1

NCS, TLS and anisotropic b-factors
by Morten Groftehauge 08 May '12

08 May '12

Hi Phenixbb, Just one more stupid question: On the phenix website it says under *current limitations of phenix.refine*"Atoms with anisotropic ADP in NCS groups". I take this to mean that phenix.refine doesn't handle the combination of TLS and NCS well, is this correct? So what is the correct way of handling TLS and NCS? Do one run where individual, isotropic ADP+xyz and NCS are refined and then another where individual, isotropic ADP+xyz and TLS are refined? Or one run where individual, isotropic ADP+xyz and NCS are refined and then another where individual, isotropic ADP and TLS are refined? Or does the issues go away if you only use torsional NCS? Or adjusting the B-factor weight of NCS selections? Should B-factor weight then be set up or down? Okay, that was more than one question but at least they're all on the same topic. Cheers, Morten -- Morten K Grøftehauge, PhD Pohl Group Durham University

4 5

UPDATE-- Difficult dataset and refinement--P422? I422?Twinning?
by Zhou, Tongqing (NIH/VRC) [E] 08 May '12

08 May '12

Hi Jürgen, Kelly and Nat, I tried XDS and MosFilm, they simply cannot index the data. I reprocessed the data yesterday with HKL2000 starting from frame #30 (splitting 1-29, 30-360), I did get it indexed to I4 and refined to very nice x/y chi^2 and covering most spots, I then used the index/refinement numbers for the whole data set. it seems that other spots were from a "nicely" positioned secondary crystal which fooled the program to pick P4 space group. I scaled the data as I4 and I422 to see which one is better. The I4 data set, ~95% complete overall, 9.2% R-sym. Xtriage suggested it should have higher symmetry I422 or twin (-k, -h, -l, almost 50%). MR sent well and best SG is I4, 2 mol/ASU, r and r-free went to about 27% and 32%, but if I used the twin-law in phenix.refine and ask phenix to generate r-free-flags with twinning in mind, the numbers went to ~22% and 27%. Shell Lower Upper Average Average Norm. Linear Square limit Angstrom I error stat. Chi**2 R-fac R-fac 50.00 7.32 440.7 8.8 5.1 1.481 0.027 0.030 7.32 5.81 167.2 4.2 3.2 1.868 0.045 0.048 5.81 5.08 196.3 5.1 3.9 2.130 0.053 0.058 5.08 4.62 243.1 6.2 4.7 2.264 0.056 0.061 4.62 4.29 236.7 6.4 4.9 2.057 0.055 0.055 4.29 4.03 167.6 5.6 4.7 1.819 0.067 0.066 4.03 3.83 139.5 5.6 5.0 1.740 0.083 0.075 3.83 3.66 110.2 5.5 5.1 1.645 0.102 0.091 3.66 3.52 92.5 5.6 5.3 1.459 0.117 0.101 3.52 3.40 76.5 5.7 5.4 1.372 0.143 0.124 3.40 3.30 59.3 5.7 5.5 1.278 0.179 0.146 3.30 3.20 47.3 5.7 5.6 1.161 0.212 0.179 3.20 3.12 40.7 5.8 5.7 1.103 0.241 0.201 3.12 3.04 31.6 5.8 5.8 1.057 0.298 0.254 3.04 2.97 28.9 5.9 5.9 1.031 0.331 0.271 2.97 2.91 22.6 5.9 5.9 0.985 0.400 0.340 2.91 2.85 19.0 6.0 6.0 0.991 0.484 0.405 2.85 2.80 17.1 6.1 6.1 0.957 0.518 0.437 2.80 2.75 14.6 6.2 6.2 0.908 0.577 0.485 2.75 2.70 13.0 6.4 6.3 0.931 0.645 0.550 All reflections 112.9 5.9 5.3 1.455 0.092 0.066 The I422 data (below), 1 mol/ASU, r and r-free were about 27% and 32%. Shell Lower Upper Average Average Norm. Linear Square limit Angstrom I error stat. Chi**2 R-fac R-fac 50.00 7.32 430.0 8.4 5.0 1.344 0.025 0.028 7.32 5.81 163.4 4.0 3.0 1.489 0.040 0.042 5.81 5.08 192.4 4.8 3.6 1.694 0.046 0.052 5.08 4.62 231.6 5.6 4.3 1.790 0.049 0.053 4.62 4.29 227.8 5.8 4.5 1.534 0.045 0.043 4.29 4.03 161.3 5.0 4.2 1.401 0.055 0.052 4.03 3.83 135.3 5.0 4.4 1.367 0.066 0.059 3.83 3.66 106.2 4.9 4.4 1.302 0.082 0.071 3.66 3.52 88.1 4.8 4.5 1.253 0.097 0.082 3.52 3.40 75.0 4.9 4.6 1.193 0.113 0.096 3.40 3.30 56.7 4.8 4.6 1.121 0.144 0.115 3.30 3.20 44.8 4.8 4.7 1.061 0.172 0.141 3.20 3.12 38.0 4.8 4.7 1.011 0.203 0.166 3.12 3.04 29.1 4.7 4.7 1.029 0.253 0.209 3.04 2.97 27.7 4.8 4.8 0.985 0.261 0.209 2.97 2.91 21.4 4.9 4.8 0.917 0.320 0.266 2.91 2.85 18.3 4.9 4.8 0.949 0.380 0.316 2.85 2.80 16.1 4.9 4.9 0.955 0.431 0.358 2.80 2.75 14.2 5.0 5.0 0.940 0.483 0.399 2.75 2.70 12.6 5.2 5.2 0.886 0.536 0.462 All reflections 110.9 5.1 4.5 1.236 0.076 0.054 So my questions are: 1. Is it really twinned since including twin-law in refinment gave much better numbers? 2. When refining with twin-law, Are the maps by phenix.refine output already "corrected" mtz/map? 3. If I use xray_data.r_free_flags.generate=True, can I write the data mtz out with the generated r-free-flags so that each round of refinement uses the same CV set? Thnaks, Tongqing Tongqing Zhou, Ph.D. Staff Scientist Structural Biology Section Vaccine Research Center, NIAID/NIH Building 40, Room 4609B 40 Convent Drive, MSC3027 Bethesda, MD 20892 (301) 594-8710 (Tel) (301) 793-0794 (Cell) (301) 480-2658 (Fax) ****************************************************************************** The information in this e-mail and any of its attachments is confidential and may contain sensitive information. It should not be used by anyone who is not the original intended recipient. If you have received this e-mail in error please inform the sender and delete it from your mailbox or any other storage devices. National Institute of Allergy and Infectious Diseases shall not accept liability for any statements made that are sender's own and not expressly made on behalf of the NIAID by one of its representatives. ****************************************************************** From: Bosch, Juergen [mailto:[email protected]] Sent: Monday, May 07, 2012 1:34 PM To: PHENIX user mailing list Subject: Re: [phenixbb] Difficult dataset and refinement--P422? I422?Twinning? Why don't you index over the whole dataset (XDS) ? Or at least over a couple of images (Mosflm) Jürgen On May 7, 2012, at 12:50 PM, Kelly Daughtry wrote: I agree. The fact that two "crystals forms" appear, with related unit cells: P422 data: unit_cell = 191.6999969 191.6999969 311.0539856 90 90 90 space_group = "P 4 2 2" I422 data: unit_cell = 191.783 191.783 103.775 90 90 90 space_group = "I 4 2 2" indicated to me you have mis-indexed the P422 data. It is likely I422. Kelly ******************************************************* Kelly Daughtry, Ph.D. Post-Doctoral Fellow, Raetz Lab Biochemistry Department Duke University Alex H. Sands, Jr. Building 303 Research Drive RM 250 Durham, NC 27710 P: 919-684-5178 ******************************************************* On Mon, May 7, 2012 at 12:18 PM, Nathaniel Echols <nechols(a)lbl.gov<mailto:[email protected]>> wrote: On Mon, May 7, 2012 at 8:46 AM, Kelly Daughtry <kddaught(a)bu.edu<mailto:[email protected]>> wrote: Looking at the P422 data, it look like you have pseudo translational symmetry. Did you try processing this data as I422? >From the P422_xtriage.log: The full list of Patterson peaks is: x y z height p-value(height) ( 0.500, 0.500, 0.165 ) : 78.635 (6.634e-07) ( 0.000, 0.000, 0.330 ) : 56.292 (2.795e-05) ( 0.500, 0.500, 0.497 ) : 30.920 (1.207e-03) One word of caution: an exceptionally high off-origin peak can mean that the unit cell was measured too large, and you've integrated extra reflections that are really non-existent. (I'm not sure what the threshold for this is, but 80% seems pretty large.) Splitting and various indexing artifacts can sometimes lead to this. I'd recommend running labelit.index on the images and seeing what it thinks the lattice should be. -Nat _______________________________________________ phenixbb mailing list phenixbb(a)phenix-online.org<mailto:[email protected]> http://phenix-online.org/mailman/listinfo/phenixbb _______________________________________________ phenixbb mailing list phenixbb(a)phenix-online.org<mailto:[email protected]> http://phenix-online.org/mailman/listinfo/phenixbb ...................... Jürgen Bosch Johns Hopkins University Bloomberg School of Public Health Department of Biochemistry & Molecular Biology Johns Hopkins Malaria Research Institute 615 North Wolfe Street, W8708 Baltimore, MD 21205 Office: +1-410-614-4742 Lab: +1-410-614-4894 Fax: +1-410-955-2926 http://web.mac.com/bosch_lab/

5 5

redundancy of merged multiple datasets
by CPMAS Chen 08 May '12

08 May '12

Dear all, I have merged multiple datasets obtained from the same crystals, different spots of course, it certainly increases the signal/noise ratio and makes the resolution go higher a little bitter. But, the con is, the redundancy is increased to ~20 after I merged 6 datasets. Whereas the individual dataset has the redundancy around 3~5. Is this acceptable? Thanks! *************************************************** CPMAS Chen Research Associate University of Pittsburgh School of Medicine Department of Anesthesiology ******************************************************

3 2

Re: [phenixbb] UPDATE-- Difficult dataset and refinement--P422? I422?Twinning?
by Kay Diederichs 08 May '12

08 May '12

> On May 8, 2012, at 11:11 AM, Zhou, Tongqing (NIH/VRC) [E] wrote: > > Hi J?rgen, Kelly and Nat, > > I tried XDS and MosFilm, they simply cannot index the data. > Hi, "simply cannot index" is hard to believe. 90% of indexing failures are due to wrong header values, 9% to the spindle rotating in the "wrong" direction, and the last 1% has other causes. I have yet to see a single-crystal dataset that cannot be indexed. If you are willing to confidentially share (at least some of) your data, I could try to process them with XDS - at least you get another opionion. Easiest way: copy them into a Dropbox or Wuala folder and share that. best, Kay

2 1

phenix.refine crashes
by Jianghai Zhu 08 May '12

08 May '12

Hi, I am using phenix.refine dev-1051. When I was trying flip_peptide, it crashes with the following error messages. Validate: |--------START--------| |--------FINAL--------| residue map_cc 2mFo-DFc mFo-DFc map_cc 2mFo-DFc mFo-DFc Traceback (most recent call last): File "/home/jzhu/phenix/phenix-dev-1051/build/intel-linux-2.6-x86_64/../../phenix/phenix/command_line/refine.py", line 11, in <module> command_line.run(command_name="phenix.refine", args=sys.argv[1:]) File "/home/jzhu/phenix/phenix-dev-1051/phenix/phenix/refinement/command_line.py", line 146, in run call_back_handler=call_back_handler) File "/home/jzhu/phenix/phenix-dev-1051/phenix/phenix/refinement/driver.py", line 1499, in run ncs_manager = self.ncs_manager) File "/home/jzhu/phenix/phenix-dev-1051/phenix/phenix/refinement/strategies.py", line 928, in refinement_machine params = params.flip_peptides) File "/home/jzhu/phenix/phenix-dev-1051/cctbx_project/mmtbx/refinement/flip_peptides.py", line 445, in run log=log) File "/home/jzhu/phenix/phenix-dev-1051/cctbx_project/mmtbx/refinement/fit_rotamers.py", line 749, in validate_changes sites_cart=sites_cart) File "/home/jzhu/phenix/phenix-dev-1051/cctbx_project/mmtbx/refinement/flip_peptides.py", line 152, in __call__ return (self.rama_lookup.evaluate_sites(res_type,i_seqs,sites_cart)<0.002) File "/home/jzhu/phenix/phenix-dev-1051/cctbx_project/mmtbx/rotamer/ramachandran_eval.py", line 76, in evaluate_sites "isoleucine or valine"]) AssertionError -- Jianghai

2 1

Re: [phenixbb] phenixbb Digest, Vol 78, Issue 9
by CPMAS Chen 08 May '12

08 May '12

Thank you all. Actually, as Pavel said there is a easy way to do so, I changed the cif file of the ligand, where I had restraints on the dihedral modified. Message: 3 > Date: Sat, 05 May 2012 13:48:06 -0700 > From: Pavel Afonine <pafonine(a)lbl.gov> > To: phenixbb(a)phenix-online.org > Subject: Re: [phenixbb] how to define a molecule in two rigid_body > Message-ID: <4FA59206.7090207(a)lbl.gov> > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > > Hi Charles, > > I think you are trying to achieve two goals, "keep conformation you > like" and "make sure it fits given density", while one of the two > contradicts to the other. That is either the density you are trying to > fit in your molecule or your expectations about the geometry of the > molecule (or both) are not quite right. > > I think this is how much one can tell given the information you provided. > > I don't think you need to do that sophisticated rigid-body kind of > refinement that you described in your original post. The solution must > be easier. May be if you send me the model and data and tell which part > of the model you think is in trouble then probably I could help more. > > Pavel > > On 5/5/12 10:22 AM, Nathaniel Echols wrote: > > On Sat, May 5, 2012 at 6:39 AM, CPMAS Chen<cpmasmit(a)gmail.com> wrote: > >> I am almost at final stage. but the refinement always shows error > density at > >> the single bond connecting the two rings. I have tried manually change > the > >> orientation of these two rings, but there are two concerns that I > thought > >> why I did not get the perfect fitting. > >> 1. I might not choose the perfect the orientation of these rings, and > these > >> give me a non-perfect fitting, and high b-factor. > >> 2. since phenix always tries to change the conformation a little bit to > fit > >> the density, so this also might cause some errors, or not the > conformation I > >> like. > >> > >> Then, my sense is that, if I can somehow define the two parts of the > >> molecule as two rigid bodies, then phenix only can has free rotation of > >> these two rings, and the fitting might come out good. > > You can define custom planarity restraints for the rings - there isn't > > good documentation for this, but here is what to do: > > > > 1. If you're using the command line, look here in the .eff file: > > > > refinement { > > geometry_restraints { > > edits { > > planarity { > > atom_selection = resseq 1001 and (name C1 or name C2 or name > > C3 or name C4 or name C5 or name C6) > > sigma = 0.1 > > } > > } > > } > > } > > > > (Other parameters omitted for clarify, of course. I'm not sure what > > an appropriate sigma would be - you'll probably have to experiment.) > > > > 2. If you're using the GUI, make sure you have the latest nightly > > build, and on the middle configuration tab, click "Custom geometry > > restraints", which will open a window that includes a tab for creating > > planarity restraints. > > > > Technically, the two planes won't actually be "rigid bodies" (i.e. > > "constrained"), but if you make the planarity restraint sufficiently > > tight there will be very little deviation from ideality. (Just don't > > make it too tight, or you'll confuse the minimizer.) > > > > -Nat > > _______________________________________________ > > phenixbb mailing list > > phenixbb(a)phenix-online.org > > http://phenix-online.org/mailman/listinfo/phenixbb > > > > ------------------------------ > > -- *************************************************** CPMAS Chen Research Associate University of Pittsburgh School of Medicine Department of Anesthesiology ******************************************************

1 0

Difficult dataset and refinement--P422? I422?Twinning?
by Zhou, Tongqing (NIH/VRC) [E] 07 May '12

07 May '12

Dear All, I have collected several dataset for an antibody:antigen complex. It apparently indexed to P4 space group with large cell dimensions of 191.7000/191.7000/311.0540. When I processed as P1 and ran Pointless, it suggested P422 as space group. P4212 gives the best MR solution with 6 mol/ASU (18 chains total ). However, ridigbody and subsequent refinement with TLS could not reduce the r and r-free which stayed at ~48%. Tried to go P4, P222...(12 mol/ASU) with twinning, but the refinement failed due to insufficient computer memory (12 GB), it was also impossible to refine in P1 due to the same reason. But for some data sets, HKL2000 also gave I4 space group with c axis reduced to 1/3 of that of P4. With the I4 data (~2.7 A resolution), I422 was the best MR space group (1 mol/ASU), the refinement went down to r=29% and r-free=34%, but won't go down further. Please advise if it's possible to use the data Tongqing Tongqing Zhou, Ph.D. Staff Scientist Structural Biology Section Vaccine Research Center, NIAID/NIH Building 40, Room 4609B 40 Convent Drive, MSC3027 Bethesda, MD 20892 (301) 594-8710 (Tel) (301) 793-0794 (Cell) (301) 480-2658 (Fax) ****************************************************************************** The information in this e-mail and any of its attachments is confidential and may contain sensitive information. It should not be used by anyone who is not the original intended recipient. If you have received this e-mail in error please inform the sender and delete it from your mailbox or any other storage devices. National Institute of Allergy and Infectious Diseases shall not accept liability for any statements made that are sender's own and not expressly made on behalf of the NIAID by one of its representatives. ******************************************************************

4 5

how to define a molecule in two rigid_body
by CPMAS Chen 05 May '12

05 May '12

Thanks, Ralf. I am almost at final stage. but the refinement always shows error density at the single bond connecting the two rings. I have tried manually change the orientation of these two rings, but there are two concerns that I thought why I did not get the perfect fitting. 1. I might not choose the perfect the orientation of these rings, and these give me a non-perfect fitting, and high b-factor. 2. since phenix always tries to change the conformation a little bit to fit the density, so this also might cause some errors, or not the conformation I like. Then, my sense is that, if I can somehow define the two parts of the molecule as two rigid bodies, then phenix only can has free rotation of these two rings, and the fitting might come out good. Charles %%%%%%%%%%%%%%%%%%%%% University of Pittsburgh School of Medicine Department of Aneshesiology %%%%%%%%%%%%%%%%%%%%%

3 2

how to define a molecule in two rigid_body
by CPMAS Chen 05 May '12

05 May '12

Dear Colleagues, First, I would say sorry to bother your guys at Friday evening. :) I have a molecule which has two rings connected by a single bond. I would like this two rings freely rotate around the single bond, but keep each ring conformation fixed (as there are sidechains connected to each ring). Could you help me do this rigid_body definition, or suggest a better idea to do so in phenix? Regards and Thanks! Charles Chen. %%%%%%%%%%%%%%%%%%%%% University of Pittsburgh School of Medicine Department of Aneshesiology %%%%%%%%%%%%%%%%%%%%%

2 1

geometry weight optimization with mlhl target
by Mooers, Blaine H.M. (HSC) 04 May '12

04 May '12

Dear Phenixbb, How does one optimize geometry when using the mlhl target in refinement with Henderson-Lattman coefficients in the input reflection file? The wxc_scale keyword is not recognized when I used the mlhl target, but it is recognized when I use the ml target by setting main.use_experimental_phases=false. Blaine Blaine Mooers Assistant Professor Department of Biochemistry and Molecular Biology University of Oklahoma Health Sciences Center S.L. Young Biomedical Research Center Rm. 466 Letter address: Shipping address: P.O. Box 26901, BRC 466 975 NE 10th Street, BRC 466 Oklahoma City, OK 73190 Oklahoma City, OK 73104-5419 office: (405) 271-8300 lab: (405) 271-8313 fax: (405) 271-3910 e-mail: blaine-mooers(a)ouhsc.edu<mailto:[email protected]> webpage: http://www.oumedicine.com/department-of-biochemistry-and-molecular-biology/…

2 2

secondary structure restraints helices/helixes proline
by Morten Groftehauge 04 May '12

04 May '12

Hi guys, How does phenix.refine handle prolines in helices? When I look in high resolution structures it seems that the carbonyl of the residue four positions before the proline (also in the helix) should be more perpendicular than parallel to the helix. This makes sense since there isn't any nitrogen for it to hydrogen bond to. But when I refine my structure it seems as if phenix.refine is trying to push the carbonyl into parallel position. Am I imagining things? It's not the best data and this is a troublesome region. Cheers, Morten -- Morten K Grøftehauge, PhD Pohl Group Durham University

3 3

Re: [phenixbb] Geometry Restraints - Anisotropic truncation
by Terwilliger, Thomas C 03 May '12

03 May '12

Hi Kendall, Yes, I think you could use this kind of approach to make overall decisions of any kind, including those you suggest. I would not use Rsleep for anything at all, other than calculating a final number. I would use a fixed Rfree set (which could be a subset of the total free set or the whole set) for all such decision making. If a lot of such decisions are made with Rfree...then yes it would be good to have an Rsleep to make sure that everything is ok. All the best, Tom T ________________________________ From: Kendall Nettles [knettles(a)scripps.edu] Sent: Thursday, May 03, 2012 9:05 AM To: Terwilliger, Thomas C; PHENIX user mailing list Subject: Fwd: [phenixbb] Geometry Restraints - Anisotropic truncation Hi Tom, Do you think something like this could be used during refinement to identify the "best" resolution limits? If you have an Rsleep set would Rfree be sufficient for this? I imagine collecting data with a ring of noise and then let the optimal resolution be determined during refinement. My understanding of this is that the modern refinement algorithms can handle some noise in the reflections, but maybe this could be a way to optimize how much signal is needed to contribute in a positive fashion? Kendall

6 7

Re: [phenixbb] Geometry Restraints - Anisotropic truncation
by Leonid Sazanov 03 May '12

03 May '12

Hi, so will we see now the possibility to apply anisotropic resolution limits truncation in Phenix? Despite the fact that a lot of people use UCLA server, there is still no possibility to do this in any major crystallographic software, AFAIK. Phaser does the scaling, but no truncation, and noise sitting in those regions where there is no signal, is a real culprit. Large SIGFs there do not resolve the problem completely. Thanks. >>>>>> Message: 4 Date: Wed, 25 Apr 2012 09:05:57 -0400 From: Mario Sanches <mariosan(a)gmail.com> To: PHENIX user mailing list <phenixbb(a)phenix-online.org> Subject: [phenixbb] Geometry Restraints - UPDATE .......... It turns out that the reason to my nightmares was a highly anisotropic dataset. When I tried to fix geometry related problems in coot, phenix.refine was twisting the geometry back again, probably because it was trying to fit a lot of noise due to the anisotropy. Everything was getting worst, geometry, clashscore, and Ramachandran. Pavel finally corrected it by using this server: http://services.mbi.ucla.edu/anisoscale/ ............ -- Dr. Leonid A. Sazanov Research group leader Medical Research Council Mitochondrial Biology Unit Wellcome Trust / MRC Building Hills Road Cambridge CB2 0XY WEB: www.mrc-mbu.cam.ac.uk Tel: +44-1223-252910 Fax: +44-1223-252915

13 32

Secondary Structure Definitions
by Antony Oliver 03 May '12

03 May '12

Dear all, Is there a phenix jiffy to assign (and write!) secondary structure definitions to a pdb file? I'm currently using MacPyMOL to put together some figures, and I can see that some of the secondary structure elements in the cartoon representation aren't being drawn correctly. Being slightly old-school, I used to use DSSP and the little program dssp2pdb to extract the information and put it into the PDB header. Is there a simpler / easier way to do this? (NB: I'm aware that CCP4MG copes better with SS assignment, but I prefer to use PyMOL in this instance). With thanks, Tony. --- Dr Antony W Oliver Senior Research Fellow CR-UK DNA Repair Enzymes Group Genome Damage and Stability Centre Science Park Road University of Sussex Falmer, Brighton, BN1 9RQ email: antony.oliver(a)sussex.ac.uk<mailto:[email protected]> tel (office): +44 (0)1273 678349 tel (lab): +44 (0)1273 677512

3 5

Fwd: Geometry Restraints - Anisotropic truncation
by Kendall Nettles 03 May '12

03 May '12

Hi Tom, Do you think something like this could be used during refinement to identify the "best" resolution limits? If you have an Rsleep set would Rfree be sufficient for this? I imagine collecting data with a ring of noise and then let the optimal resolution be determined during refinement. My understanding of this is that the modern refinement algorithms can handle some noise in the reflections, but maybe this could be a way to optimize how much signal is needed to contribute in a positive fashion? Kendall From: "Terwilliger, Thomas C" <terwilliger(a)lanl.gov<mailto:[email protected]>> Date: May 3, 2012 10:23:29 AM EDT To: PHENIX user mailing list <phenixbb(a)phenix-online.org<mailto:[email protected]>> Subject: Re: [phenixbb] Geometry Restraints - Anisotropic truncation Reply-To: PHENIX user mailing list <phenixbb(a)phenix-online.org<mailto:[email protected]>> Hi Kendall, This could work. You could define a fixed set of test reflections, and never touch these, and never include them in refinement, and always use this fixed set to calculate a free R. Then you could do whatever you want, throw away some work reflections, etc, refine, and evaluate how things are working with the fixed free R set. All the best, Tom T ________________________________________ From: phenixbb-bounces(a)phenix-online.org<mailto:[email protected]> [phenixbb-bounces(a)phenix-online.org<mailto:[email protected]>] on behalf of Kendall Nettles [knettles(a)scripps.edu<mailto:[email protected]>] Sent: Thursday, May 03, 2012 7:05 AM To: PHENIX user mailing list Subject: Re: [phenixbb] Geometry Restraints - Anisotropic truncation Hi Pavel, Could you use a similar approach to figuring out where to cut your data in general? Could you compare the effects of throwing out reflections in different bins, based on I/sigma, for example, and use this to determine what is truly noise? I might predict that as you throw out "noise" reflections you will see a larger drop in Rfree than from throwing out "signal" reflections, which should converge as you approach the "true" resolution. While we don't use I/sigma exclusively, we do to tend towards cutting most of our data sets at the same i/sigma, around 1.5. It would be great if there was a more scientific approach. Best, Kendall On May 3, 2012, at 7:45 AM, Pavel Afonine wrote: Hi Kendall, removing same amount of data randomly gives Rwork/Rfree ~ 30/35%. Pavel On 5/3/12 4:13 AM, Kendall Nettles wrote: Hi Pavel, What happens if you throw out that many reflections that have signal? Can you take out a random set of the same size? Best, Kendall On May 3, 2012, at 2:41 AM, "Pavel Afonine"<pafonine(a)lbl.gov<mailto:[email protected]>> wrote: Hi Kendall, I just did this quick test: calculated R-factors using original and anisotropy-corrected Mike Sawaya's data (*) Original: r_work : 0.3026 r_free : 0.3591 number of reflections: 26944 Truncated: r_work : 0.2640 r_free : 0.3178 number of reflections: 18176 The difference in R-factors is not too surprising given how many reflections was removed (about 33%). Pavel (*) Note, the data available in PDB is anisotropy corrected. The original data set was kindly provided to me by the author. On 5/2/12 5:25 AM, Kendall Nettles wrote: I didnt think the structure was publishable with Rfree of 33% because I was expecting the reviewers to complain. We have tested a number of data sets on the UCLA server and it usually doesn't make much difference. I wouldn't expect truncation alone to change Rfree by 5%, and it usually doesn't. The two times I have seen dramatic impacts on the maps ( and Rfree ), the highly anisotrophic sets showed strong waves of difference density as well, which was fixed by throwing out the noise. We have moved to using loose data cutoffs for most structures, but I do think anisotropic truncation can be helpful in rare cases. Kendall On May 1, 2012, at 3:07 PM, "Dale Tronrud"<det102(a)uoxray.uoregon.edu<mailto:[email protected]>> wrote: While philosophically I see no difference between a spherical resolution cutoff and an elliptical one, a drop in the free R can't be the justification for the switch. A model cannot be made more "publishable" simply by discarding data. We have a whole bunch of empirical guides for judging the quality of this and that in our field. We determine the resolution limit of a data set (and imposing a "limit" is another empirical choice made) based on Rmrg, or Rmes, or Rpim getting too big or I/sigI getting too small and there is no agreement on how "too big/small" is too "too big/small". We then have other empirical guides for judging the quality of the models we produce (e.g. Rwork, Rfree, rmsds of various sorts). Most people seem to recognize that the these criteria need to be applied differently for different resolutions. A lower resolution model is allowed a higher Rfree, for example. Isn't is also true that a model refined to data with a cutoff of I/sigI of 1 would be expected to have a free R higher than a model refined to data with a cutoff of 2? Surely we cannot say that the decrease in free R that results from changing the cutoff criteria from 1 to 2 reflects an improved model. It is the same model after all. Sometimes this shifting application of empirical criteria enhances the adoption of new technology. Certainly the TLS parametrization of atomic motion has been widely accepted because it results in lower working and free Rs. I've seen it knock 3 to 5 percent off, and while that certainly means that the model fits the data better, I'm not sure that the quality of the hydrogen bond distances, van der Waals distances, or maps are any better. The latter details are what I really look for in a model. On the other hand, there has been good evidence through the years that there is useful information in the data beyond an I/sigI of 2 or an Rmeas> 100% but getting people to use this data has been a hard slog. The reason for this reluctance is that the R values of the resulting models are higher. Of course they are higher! That does not mean the models are of poorer quality, only that data with lower signal/noise has been used that was discarded in the models you used to develop your "gut feeling" for the meaning of R. When you change your criteria for selecting data you have to discard your old notions about the acceptable values of empirical quality measures. You either have to normalize your measure, as Phil Jeffrey recommends, by ensuring that you calculate your R's with the same reflections, or by making objective measures of map quality. Dale Tronrud P.S. It is entirely possible that refining a model to a very optimistic resolution cutoff and calculating the map to a lower resolution might be better than throwing out the data altogether. On 5/1/2012 10:34 AM, Kendall Nettles wrote: I have seen dramatic improvements in maps and behavior during refinement following use of the UCLA anisotropy server in two different cases. For one of them the Rfree went from 33% to 28%. I don't think it would have been publishable otherwise. Kendall On May 1, 2012, at 11:10 AM, Bryan Lepore wrote: On Mon, Apr 30, 2012 at 4:22 AM, Phil Evans<pre(a)mrc-lmb.cam.ac.uk<mailto:[email protected]>> wrote: Are anisotropic cutoff desirable? is there a peer-reviewed publication - perhaps from Acta Crystallographica - which describes precisely why scaling or refinement programs are inadequate to ameliorate the problem of anisotropy, and argues why the method applied in Strong, et. al. 2006 satisfies this need? -Bryan _______________________________________________ phenixbb mailing list phenixbb(a)phenix-online.org<mailto:[email protected]> http://phenix-online.org/mailman/listinfo/phenixbb _______________________________________________ phenixbb mailing list phenixbb(a)phenix-online.org<mailto:[email protected]> http://phenix-online.org/mailman/listinfo/phenixbb _______________________________________________ phenixbb mailing list phenixbb(a)phenix-online.org<mailto:[email protected]> http://phenix-online.org/mailman/listinfo/phenixbb _______________________________________________ phenixbb mailing list phenixbb(a)phenix-online.org<mailto:[email protected]> http://phenix-online.org/mailman/listinfo/phenixbb _______________________________________________ phenixbb mailing list phenixbb(a)phenix-online.org<mailto:[email protected]> http://phenix-online.org/mailman/listinfo/phenixbb

1 0

How to include restraints when refining DNA?
by Xun Lu 03 May '12

03 May '12

Dear all, I am refining a protein-DNA complex structure at 2.9A resolution. The R/Rfree is 0.2181/0.2501, which is pretty good, but I can tell that the DNA is not well refined. Some sugar puckers are not right. The distances between some base-paired bases are kind of short (~2.5A). I've tried including a reference model for the DNA, or including a cif file, and I've also tried to turn on the secondary_structure_restraints.... nothing worked. I must have not done it in a correct way. But how to do it? And how to deal with sugar puckers? In refmac, I can set the distances between bases during the refinement, so there must be a way to do it in phenix as well? Any advice is welcome. Thanks! Xun -- Department of Molecular and Structural Biochemistry North Carolina State University

4 3

anisotropic induced noise amplification
by James Thompson 02 May '12

02 May '12

Anyone willing to summarize Xtriage's algorithm that determines whether "anisotropic induced noise amplification" is present in diffraction data? What is this noise amplification? Is it anisotropic correction factors that over-fit the data thereby inducing noise? Many thanks, Jim T ---------------- Anisotropy analyses ---------------- Anisotropy ( [MaxAnisoB-MinAnisoB]/[MaxAnisoB] ) : 3.596e-01 Anisotropic ratio p-value : 0.000e+00 The p-value is a measure of the severity of anisotropy as observed in the PDB. The p-value of 0.000e+00 indicates that roughly 100.0 % of datasets available in the PDB have an anisotropy equal to or worse than this dataset. For the resolution shell spanning between 2.36 - 2.20 Angstrom, the mean I/sigI is equal to 3.37. 42.2 % of these intensities have an I/sigI > 3. When sorting these intensities by their anisotropic correction factor and analysing the I/sigI behavior for this ordered list, we can gauge the presence of 'anisotropy induced noise amplification' in reciprocal space. The quarter of Intensities *least* affected by the anisotropy correction show <I/sigI> : 3.82e+00 Fraction of I/sigI > 3 : 5.01e-01 ( Z = 3.59 ) The quarter of Intensities *most* affected by the anisotropy correction show <I/sigI> : 2.02e+00 Fraction of I/sigI > 3 : 1.84e-01 ( Z = 10.78 ) The combined Z-score of 11.36 indicates that there probably is significant systematic noise amplification that could possibly lead to artefacts in the maps or difficulties in refinement

1 0

Re: [phenixbb] Scaling of diffraction data
by jens j birktoft 02 May '12

02 May '12

Hi Everybody, Is there a protocol for re-scaling several sets of diffraction data against a common reference set, either native or based on a Fcalc set computed from a common input model. The are from a number of derivatives, are all isomorphus, and are anomalous data . We want to compare the relative peak heights in Fo-Fc and Anom. difference maps Cheers Jens +++++++++++++++++++++++++++++++++++++++++++++++++++ Jens J. Birktoft e-mail: birktoft @ nyu.edu <jens.knold(a)gmail.com> ; Phone: 212-749-5057 +++++++++++++++++++++++++++++++++++++++++++++++++++

5 4

Xtriage problem
by Abhinav Kumar 02 May '12

02 May '12

Hi, When I try to do R vs R analysis for calculated data using xtriage, it has a problem accepting calc labels. source /home/sw/rhel5/x86_64/phenix/phenix-1.7.3-928/phenix_env phenix.xtriage_1.7.3-928 junk.mtz calc_labels=FC ==> Sorry: No matching array: scaling.input.xray_data.calc_labels=FC Howevere, it works fine with an earlier version: source /home/sw/rhel5/x86_64/phenix/phenix-1.6.4-486/phenix_env phenix.xtriage_1.6.4-486 junk.mtz calc_labels=FC Thanks, Abhinav JCSG@SSRL, SLAC (650) 926-2992 -------- Original Message -------- Subject: RE: SP16569N-180314 Date: Tue, 1 May 2012 15:09:00 -0700 From: Miller, Mitchell D. <mmiller(a)slac.stanford.edu> To: Kumar, Abhinav <abhinavk(a)slac.stanford.edu> Hmm, it seems like it may be a bug introduced into phenix. (or else they changed they syntax and did not update the documentation) Can you report it? I get it to do the RvR analysis for calculated data if I use source /home/sw/rhel5/x86_64/phenix/phenix-1.6.4-486/phenix_env phenix.xtriage_1.6.4-486 junk.mtz calc_labels=FC but if I use source /home/sw/rhel5/x86_64/phenix/phenix-1.7.1-743/phenix_env phenix.xtriage_1.7.1-743 junk.mtz calc_labels=FC or source /home/sw/rhel5/x86_64/phenix/phenix-1.7.3-928/phenix_env phenix.xtriage_1.7.3-928 junk.mtz calc_labels=FC then I get the error Sorry: No matching array: scaling.input.xray_data.calc_labels=FC Thanks, Mitch P.S. the results I got were: R vs R statistic: R_abs_twin =<|I1-I2|>/<|I1+I2|> Lebedev, Vagin, Murshudov. Acta Cryst. (2006). D62, 83-95 R_abs_twin observed data : 0.451 R_abs_twin calculated data : 0.502 R_sq_twin =<(I1-I2)^2>/<(I1+I2)^2> R_sq_twin observed data : 0.279 R_sq_twin calculated data : 0.335 The analyses of the Patterson function reveals a significant off-origin peak that is 44.82 % of the origin peak, indicating pseudo translational symmetry. The chance of finding a peak of this or larger height by random in a structure without pseudo translational symmetry is equal to the 1.4447e-04. See the relevant section of the logfile for more details. The results of the L-test indicate that the intensity statistics behave as expected. No twinning is suspected. Even though no twinning is suspected, it might be worthwhile carrying out a refinement using a dedicated twin target anyway, as twinned structures with low twin fractions are difficult to distinguish from non-twinned structures. -----Original Message----- From: Kumar, Abhinav Sent: Tuesday, May 01, 2012 1:30 PM To: Miller, Mitchell D. Subject: Re: SP16569N-180314 I am having trouble inputting Calc labels to xtriage: My mtz file has the following columns: FreeR_flag FP SIGFP FC PHIC FC_ALL PHIC_ALL FWT PHWT DELFWT PHDELWT FOM PHCOMB I tried "FC,PHIC" and "FC" for calc labels but xtriage does not like it. It writes: No matching array: scaling.input.xray_data.calc_labels=FC Possible choices: /mnt/exdsstaff/abhinavk/Refine/SP16569N-180314/RESULTS/current.mtz:FP,SIGFP /mnt/exdsstaff/abhinavk/Refine/SP16569N-180314/RESULTS/current.mtz:FOM /mnt/exdsstaff/abhinavk/Refine/SP16569N-180314/RESULTS/current.mtz:PHCOMB Thanks, Abhinav JCSG@SSRL, SLAC (650) 926-2992 On 05/01/2012 12:51 PM, Miller, Mitchell D. wrote: Also, you could try running xtriage with FC's from the untwined refinement to look at the R vs R stats for Fc vs Fobs.

2 2

Re: [phenixbb] Geometry Restraints - Anisotropic truncation
by Frank von Delft 02 May '12

02 May '12

Hi Pavel > we discussed this with Randy the other day.. A couple of copy-pasts from > that discussion: > > In general, given highly anisotropic data set: > > 1) maps calculated using all (unmodified) data by phenix.refine, > phenix.maps and similar tools are better than maps calculated using > anisotropy truncated data. So, yes, for the purpose of map calculation > there is no need to do anything: Phenix map calculation tools deal with > anisotropy very well. How did you define "better"? If there are a lot of reflections without signal, that makes them essentially missing, so by including them, you're effectively filling in for those reflections with only DFc. If anisotropy is very strong (i.e. many missing reflections), does that not introduce very significant model bias? The maps would look cleaner, though. (I haven't tested myself, btw; I'm genuinely wondering.) > 2) phenix.refine refinement may fail if one uses original anisotropic > data set. This is probably because the ML target does not use > experimental sigmas (and anisotropy correction by UCLA server is nothing > but Miller index dependent removing the data by sigma criterion - yeah, > that old well criticized practice of throwing away the data that you > worked hard to measure!). May be using sigmas in ML calculation could > solve the problem but that has to be proved. If there are large swathes of reciprocal space without signal - I don't see why shouldn't that be excluded? Tossing out individual reflections is of course silly, but what's wrong with trimming down the edges of the elipsoid? Cheers phx

4 4

Inputting DNA for Phenix MR and/or Autobuild
by Charles Allerston 02 May '12

02 May '12

Dear all, I can't find much in the way of documentation for this. Can someone point me to some please? I basically want to know the correct format for entering DNA into the search for MR and inputting into Autobuild. If I have a double stranded DNA do I enter it as >DNA1 AATT:AATT >DNA1 AATT TTAA etc. ? cheers, charlie Dr. Charles Allerston Genome Integrity & Repair Structural Genomics Consortium Nuffield Department of Medicine Old Road Campus University of Oxford OX3 7DQ http://www.thesgc.org/scientists/groups/oxford/genome-integrity-and-repair

8 12

Call for submissions to June 2012 Computational Crystallography Newsletter (CCN)
by Nigel Moriarty 01 May '12

01 May '12

Calling for articles and short communications of interest to structural biologists. The deadline for publication in the June 2012 issue is 29 May, 2012. The Computational Crystallography Newsletter (CCN) is an electronic newsletter for structural biologists, and is published online every 6 months. Feature articles, meeting announcements and reports can be submitted to me at any time for consideration. Submission of text by email or word-processing files using the CCN templates is requested. Past newsletters and the template are available at www.phenix-online.org/newsletter. Cheers Nigel -- Nigel W. Moriarty Building 64R0246B, Physical Biosciences Division Lawrence Berkeley National Laboratory Berkeley, CA 94720-8235 Phone : 510-486-5709 Email : NWMoriarty(a)LBL.gov Fax : 510-486-5909 Web : CCI.LBL.gov

1 0