- phenixbb - phenix-online.org

R-factor difference between phenix and sftools
by Simon Jenni 13 Oct '14

13 Oct '14

Hi, we refined a structure in phenix.refine (dev-1810) and the R-factors are: Final R-work = 0.3221, R-free = 0.3638 phenix.cc_star agrees: phenix.cc_star com_001.pdb com_001.mtz \ f_obs_labels="F-obs,SIGF-obs" \ f_model_labels="F-model,PHIF-model" \ unmerged_data="xscale.hkl" r_work: 0.322 r_free: 0.364 so does phenix.model_vs_data: phenix.model_vs_data com_001.pdb com_001.mtz r_work : 0.3221 r_free : 0.3638 sigma_cutoff : None However, when I calculate the R-factors with sftools, I get a discrepancy between the phenix and sftools results: sftools << eof read com_001_f_model.mtz Y select col R_FREE_FLAGS > 0 correl col FOBS FMODEL select invert correl col FOBS FMODEL quit Y eof R-work: 33.6 R-free: 38.1 Does anyone perhaps know what causes this difference? With best regards, Simon PS: The structure is in spacegroup P212121 with strong pseudocentering tNCS (dimer in the asymmetric unit, native Patterson peak at 0.500 0.500 0.497). No indication for a P21 twin (L-test and refinement in P21 with twin law), not higher symmetry spacegroup I222 (alternating strong and weak reflections are visible in the diffraction patterns, resolved structural differences in the two protomers of the dimer). Two datasets from different beamlines. As might be expected, the R-factors after refinement are unusually high (e.g. 28/35 [conventional resolution cutoff at 3.1 A], 32/36 with Kay's CC cufoff at 2.6 A). I wanted to compare refinements between different programs to see how they cope with the unexpected bimodal amplitude distribution.

4 8

ligand fitting
by Sneha Rangarajan 13 Oct '14

13 Oct '14

Hello everyone, I have a question about ligand fitting into density. At this point my maps look quite good with decent density for the peptide (ligand)[Rfactprs 26/31]. I tried using ligandfit by giving it the pdb and mtz of the ligand free model along with peptide.pdb (peptide stripped from a pdb where it was complexed with a homologous protein). However the output was a ligand.pdb file with a CC of 0.49. I am not sure how to interpret this. Does this mean it could not find the density for the ligand? Is there a better way to fit the peptide into density? Thanks, S From: phenixbb-bounces(a)phenix-online.org [mailto:[email protected]] On Behalf Of Sneha Rangarajan Sent: Wednesday, October 08, 2014 10:30 AM To: Nathaniel Echols Cc: phenixbb(a)phenix-online.org Subject: Re: [phenixbb] (no subject) This was a great idea. My Rfactors after a second round of autobuild are now 25/32. I think it might be getting there afterall ☺ S From: Nathaniel Echols [mailto:[email protected]] Sent: Friday, October 03, 2014 3:08 PM To: Sneha Rangarajan Cc: Pavel Afonine; phenixbb(a)phenix-online.org<mailto:[email protected]> Subject: Re: [phenixbb] (no subject) On Fri, Oct 3, 2014 at 11:58 AM, Sneha Rangarajan <rsneha(a)umd.edu<mailto:[email protected]>> wrote: I did another round of refinement with default settings (XYZ,realsp, IndB and occ) with and without weight optimization. Without weight opt, the Rfactors are 23/36 with RMSbonds-0.0108 and RMSangles-1.750 One idea would be to run AutoBuild again. I've seen cases before where it didn't converge using the default settings, and feeding a previous result back into the program for a second run produced significantly better models. It might help get rid of the overfitting. -Nat

5 12

Re: [phenixbb] R-factor difference between phenix and sftools
by Ed Pozharski 13 Oct '14

13 Oct '14

To be fair, I made the same mistake initially, but the question is about calculating Rs with sftools using output mtz from phenix.refine, not sfcheck. Thusly, bulk solvent explanation does not seem to apply. My other theory (impossible to check without actual data) was that phenix.refine outlier reflection rejection makes the difference here, although it's surprising if effect is so noticeable. Sent on a Sprint Samsung Galaxy S® III <div>-------- Original message --------</div><div>From: Pavel Afonine <pafonine(a)lbl.gov> </div><div>Date:10/13/2014 2:21 PM (GMT-05:00) </div><div>To: Tim Gruene <tg(a)shelx.uni-ac.gwdg.de>, phenixbb(a)phenix-online.org </div><div>Subject: Re: [phenixbb] R-factor difference between phenix and sftools </div><div> </div>Thanks Tim, yes missed this one! Hi Simon, the differences in R-factors you observe are most likely because we use improved bulk-solvent modeling and overall scaling procedure. This is described here: Bulk-solvent and overall scaling revisited: faster calculations, improved results. Afonine PV, Grosse-Kunstleve RW, Adams PD, Urzhumtsev A. Acta Cryst. D69, 625-34 (2013). Specifically, Figures 3-4 in this paper make your observations very expectable. Other contributions to differences are discussed here: phenix.model_vs_data: a high-level tool for the calculation of crystallographic model and data statistics. P.V. Afonine, R.W. Grosse-Kunstleve, V.B. Chen, J.J. Headd, N.W. Moriarty, J.S. Richardson, D.C. Richardson, A. Urzhumtsev, P.H. Zwart, P.D. Adams J. Appl. Cryst. 43, 677-685 (2010). Pavel On 10/13/14 11:10 AM, Tim Gruene wrote: Dear Simon, I found this a very puzzling observation and given the high throughput of this board, it is likewise puzzling you have not received a reply yet - or have you? Best regards, Tim On 10/12/2014 08:25 PM, Simon Jenni wrote: Hi, we refined a structure in phenix.refine (dev-1810) and the R-factors are: Final R-work = 0.3221, R-free = 0.3638 phenix.cc_star agrees: phenix.cc_star com_001.pdb com_001.mtz \ f_obs_labels="F-obs,SIGF-obs" \ f_model_labels="F-model,PHIF-model" \ unmerged_data="xscale.hkl" r_work: 0.322 r_free: 0.364 so does phenix.model_vs_data: phenix.model_vs_data com_001.pdb com_001.mtz r_work : 0.3221 r_free : 0.3638 sigma_cutoff : None However, when I calculate the R-factors with sftools, I get a discrepancy between the phenix and sftools results: sftools << eof read com_001_f_model.mtz Y select col R_FREE_FLAGS > 0 correl col FOBS FMODEL select invert correl col FOBS FMODEL quit Y eof R-work: 33.6 R-free: 38.1 Does anyone perhaps know what causes this difference? With best regards, Simon PS: The structure is in spacegroup P212121 with strong pseudocentering tNCS (dimer in the asymmetric unit, native Patterson peak at 0.500 0.500 0.497). No indication for a P21 twin (L-test and refinement in P21 with twin law), not higher symmetry spacegroup I222 (alternating strong and weak reflections are visible in the diffraction patterns, resolved structural differences in the two protomers of the dimer). Two datasets from different beamlines. As might be expected, the R-factors after refinement are unusually high (e.g. 28/35 [conventional resolution cutoff at 3.1 A], 32/36 with Kay's CC cufoff at 2.6 A). I wanted to compare refinements between different programs to see how they cope with the unexpected bimodal amplitude distribution. _______________________________________________ phenixbb mailing list phenixbb(a)phenix-online.org http://phenix-online.org/mailman/listinfo/phenixbb _______________________________________________ phenixbb mailing list phenixbb(a)phenix-online.org http://phenix-online.org/mailman/listinfo/phenixbb

1 0

Refine three alternate conformations of residues and two loop confirmations.
by AMIT 13 Oct '14

13 Oct '14

Hi All, I am unable to refine three alternate conformations of a residue provided in PDB file prefixed before resname as A,B and C in Phenix dev-1370. When I make it two alternate conformations it runs OK. Also I am not able to refine two loop confirmations, the refinement somehow puts both in the major 2Fo-Fc map peak and makes the almost same, even though the starting loop has different trace.Turning off siulated annealing or making temp. lower or just doing 1 cycle is not doing any good. Any suggestions will be helpful. Thanks, Amit. ------------------------------------------------------- AMIT DAS, Ph.D. PRESENT ---- BUILDING 8600, MS-6460 BIOLOGY AND SOFT MATTER DIVISION P O BOX 2008 OAK RIDGE NATIONAL LABORATORY 1 BETHEL VALLEY ROAD OAK RIDGE, TN 37831-6460 U.S.A. PROTEIN CRYSTALLOGRAPHY SOLID STATE PHYSICS DIVISION BARC, TROMBAY, MUMBAI-400085. INDIA. --------------------------------------------------------

3 2

Test
by Paul Adams 11 Oct '14

11 Oct '14

-- Paul Adams Deputy Division Director, Physical Biosciences Division, Lawrence Berkeley Lab Division Deputy for Biosciences, Advanced Light Source, Lawrence Berkeley Lab Adjunct Professor, Department of Bioengineering, U.C. Berkeley Vice President for Technology, the Joint BioEnergy Institute Laboratory Research Manager, ENIGMA Science Focus Area Building 64, Room 248 Building 80, Room 247 Building 978, Room 4126 Tel: 1-510-486-4225, Fax: 1-510-486-5909 http://cci.lbl.gov/paul Lawrence Berkeley Laboratory 1 Cyclotron Road BLDG 64R0121 Berkeley, CA 94720, USA. Executive Assistant: Louise Benvenue [ LBenvenue(a)lbl.gov ][ 1-510-495-2506 ] --

1 0

anomalous difference map
by CPMAS Chen 10 Oct '14

10 Oct '14

Hi, phenix users, A difference map is used to identify whether there is anything not-modeled, say some ligands, ions. Then when I generate anomalous difference map, I should not put the ligand(which contains Br) in the model, right? Or in phenix, after I put the Br-ligand in the model, I should not see the anomalous difference density at the site, right? Could anyone give me a confirmation of this? Thanks! Charles -- *************************************************** Charles Chen Research Associate University of Pittsburgh School of Medicine Department of Anesthesiology ******************************************************

6 15

Re: [phenixbb] (no subject)
by Sneha Rangarajan 08 Oct '14

08 Oct '14

After running autobuild, the Rfactors were 26/39 (autobuild removed the first 30 residues from the Nter!) but the maps look much better. I ran a quick refinement with default settings and the Rfactors are now 31/39 and can now see the density for my peptide (absent in the model) coming up. I have a few questions- 1) When is a good time to start building the peptide into the density (should I wait until Rfactors improve significantly)? 2) If so, should I be masking the peptide density somehow so that further refinement doesn't flatten out that density? 3) What is a general good strategy for lowering Rfactors? (Esp if the gap between Rwork and Rfree is high..like 28/40 or so)? Thanks a lot, S -----Original Message----- From: Sneha Rangarajan Sent: Wednesday, October 01, 2014 12:54 PM To: 'Paul Adams' Cc: phenixbb(a)phenix-online.org Subject: RE: I think my problem in MR has been one of the chains. Anyway I tried the ideas you suggested and seems like they worked! I used ensembler to create a merged pdb of the problematic chain and used phaser giving both chains as multiple ensembles. It gave me a solution with LLG=474 (highest I have got so far) and TFZ=18.8. I also tried MRage and got a solution with LLG=294. In both these maps, the density for chain1 ("bad chain") is not as good as the other chain but is nonetheless far better than what I have got before. I am currently running autobuild on the solution hoping my Rfactors would be reasonable this time. Thanks a lot for your inputs. They really helped!! S -----Original Message----- From: Paul Adams [mailto:[email protected]] Sent: Tuesday, September 30, 2014 7:51 PM To: Sneha Rangarajan Subject: Re: This likes normal orthorhombic data. In this case I suggest that the most likely issue is the search model (I assume your molecule is fairly small given the unit cell dimensions). I would look at ways to create the best search model using using tools like sculptor and ensembler. You can get Phaser to search for the two domains in the same run but using multiple ensembles. You might also want to try MRage in Phenix, where you can just provide the data file and the sequence and it will try to make the best model and solve the MR problem. On Sep 30, 2014, at 3:32 PM, Sneha Rangarajan <rsneha(a)umd.edu> wrote: > Oh I see. I tried to repeat xtriage with the scala.mtz file and am attaching the log file. > > ________________________________________ > From: Sneha Rangarajan > Sent: Tuesday, September 30, 2014 5:58 PM > To: Paul Adams > Cc: help(a)phenix-online.org > Subject: RE: > > Even i thought so but after i run mosflm and process the data in p21 or p1 and run quicksymm/quickscale it always suggests that the chosen space group is not correct and that it should be p212121. > > I will try looking into Zanuda. Thanks. > > S > ________________________________________ > From: Paul Adams [pdadams(a)lbl.gov] > Sent: Tuesday, September 30, 2014 5:36 PM > To: Sneha Rangarajan > Cc: help(a)phenix-online.org > Subject: Re: > > It looks like your R-merge etc in P222 is very high, what leads you to think this is the correct point group? Can you send the xtriage output for the unmerged P2 data? > > On Sep 30, 2014, at 2:29 PM, Sneha Rangarajan <rsneha(a)umd.edu> wrote: > >> Hi Nat, >> >> I am attaching the Xtriage log file from p222.mtz. >> Would you need anything else (mtz,pdb)? >> >> Thanks, >> Sneha >> <xtriage_237.log> > > -- > Paul Adams > Deputy Division Director, Physical Biosciences Division, Lawrence > Berkeley Lab Division Deputy for Biosciences, Advanced Light Source, > Lawrence Berkeley Lab Adjunct Professor, Department of Bioengineering, > U.C. Berkeley Vice President for Technology, the Joint BioEnergy > Institute Laboratory Research Manager, ENIGMA Science Focus Area > > Building 64, Room 248 > Building 80, Room 247 > Building 978, Room 4126 > Tel: 1-510-486-4225, Fax: 1-510-486-5909 http://cci.lbl.gov/paul > > Lawrence Berkeley Laboratory > 1 Cyclotron Road > BLDG 64R0121 > Berkeley, CA 94720, USA. > > Executive Assistant: Louise Benvenue [ LBenvenue(a)lbl.gov ][ > 1-510-495-2506 ] > -- > > > > > > > > > > > > > <xtriage_264.log> -- Paul Adams Deputy Division Director, Physical Biosciences Division, Lawrence Berkeley Lab Division Deputy for Biosciences, Advanced Light Source, Lawrence Berkeley Lab Adjunct Professor, Department of Bioengineering, U.C. Berkeley Vice President for Technology, the Joint BioEnergy Institute Laboratory Research Manager, ENIGMA Science Focus Area Building 64, Room 248 Building 80, Room 247 Building 978, Room 4126 Tel: 1-510-486-4225, Fax: 1-510-486-5909 http://cci.lbl.gov/paul Lawrence Berkeley Laboratory 1 Cyclotron Road BLDG 64R0121 Berkeley, CA 94720, USA. Executive Assistant: Louise Benvenue [ LBenvenue(a)lbl.gov ][ 1-510-495-2506 ] --

3 5

PDBe workshop on programmatic access to macromolecular structure information, 20-21 Nov 2014, Cambridge UK
by swanand＠ebi.ac.uk 05 Oct '14

05 Oct '14

=============================================================== Programmatic Access to the Macromolecular Structure Information =============================================================== -------- PDBe Workshop for Programmers, 20-21 Nov 2014 -------- ============= What? The Protein Data bank in Europe (PDBe; http://pdbe.org) is organising a hands-on, interactive workshop for developers interested in using primary and value-added information about the 3D structures in PDB and EMDB. Programmatic access to macromolecular structure data as well as information about assemblies, structure quality, literature and citations, and up-to-date cross-references to other biomedical resources (UniProt, Pfam, CATH, SCOP, etc.) is a critical requirement for integrative bioinformatics analyses and workflows. PDBe has developed a set of REST APIs (http://wwwdev.ebi.ac.uk/pdbe/api/doc) that provide easy access to high-quality information regarding biomacromolecular structures archived in PDB and EMDB. They provide a simple, reliable, light-weight mechanism to query macromolecular structure data, select entries or molecules of interest, and access targeted information about them, including information about the molecules, experimental details, assemblies, motifs, function, taxonomy, structure quality, etc. The API is used by the PDBe services and is also integrated in Jmol/Jsmol. We will also present Solr/Lucene based query API for macromolecular structure information including PDB, EMDB and value added information on assemblies, structure quality and up-to-date cross-reference infromation to other biomedical databases based on SIFTS (http://pdbe.org/sifts) . ============= Is it for me? Do you (plan to) develop software or on-line resources that use or need access to information about structures in PDB or EMDB? Would you rather not maintain your private infrastructure for retrieving, parsing, storing and updating such information? If so, then this workshop is for you! In the workshop, we will explain what information is available through the API and how you can access it from your own software or resource. There is plenty of time for hands-on work where experienced PDBe developers will try to help you use the API for your specific needs. ============= When? 20 & 21 November 2014. ============= Where? EMBL-EBI (http://www.ebi.ac.uk) , Cambridge. ============= How much does it cost? We expect you to pay for (and arrange) your own travel to and from Cambridge. However, there is no fee for the workshop, and we will provide free accommodation (Holiday Inn Express; http://www.hiexpresscambridgeduxford.co.uk), transportation between the hotel and the EMBL-EBI, coffee/tea, lunches and a workshop dinner on the 20th. This is all thanks to generous funding from the Wellcome Trust to PDBe. ============= How can I apply? E-mail Swanand Gore (swanand(a)ebi.ac.uk) or Jose Dana (dana(a)ebi.ac.uk) at PDBe no later than 20 October 2014 with the following details: * a brief description of the software you have developed or are developing - i.e. what it does, in which field, how many users, relevant publications, etc. * programming language(s) you are specifically interested in (python examples and tutorials will be provided). * how you would benefit from this workshop. * any specific topics or questions you would like to see addressed in the workshop. If the workshop is oversubscribed, we will select participants based on the information provided. You will be informed by the end of October if you have been selected or not, or whether you are on the stand-by list. ============= Prerequisites? Basic programming experience in use of REST API’s. Bring your own laptop with git and iPython notebook pre-installed if you want to follow the python tutorial and examples. If you need to make plots, make sure that matplotlib is also installed. ============= Questions? Swanand Gore (swanand(a)ebi.ac.uk) or Jose Dana (dana(a)ebi.ac.uk) at PDBe will be happy to answer your queries.

1 0

Re: [phenixbb] Calculate average B-factor?
by Ed Pozharski 05 Oct '14

05 Oct '14

grep "^ATOM " filename.pdb | cut -c 61-66 | awk '{s+=$1;} END {print s/NR;}' "Nobody likes a show off, Private" Skipper Cheers Sent on a Sprint Samsung Galaxy S® III <div>-------- Original message --------</div><div>From: Chen Zhao <c.zhao(a)yale.edu> </div><div>Date:10/04/2014 4:03 PM (GMT-05:00) </div><div>To: PHENIX user mailing list <phenixbb(a)phenix-online.org> </div><div>Subject: [phenixbb] Calculate average B-factor? </div><div> </div>Dear all, I am just wondering whether there is a command line tool in phenix that calculates the average B-factor of a PDB file? Can it deal with the ANISOU records (from TLS refinement or not) properly? I looked into previous posts but the --show-adp-statistics option in phenix.pdbtools seems to be no longer available in the version (1.9-1678) I installed. Thank you so much, Chen

3 2

Calculate average B-factor?
by Chen Zhao 04 Oct '14

04 Oct '14

Dear all, I am just wondering whether there is a command line tool in phenix that calculates the average B-factor of a PDB file? Can it deal with the ANISOU records (from TLS refinement or not) properly? I looked into previous posts but the --show-adp-statistics option in phenix.pdbtools seems to be no longer available in the version (1.9-1678) I installed. Thank you so much, Chen

2 2