- phenixbb - phenix-online.org

July 2013 Computational Crystallography Newsletter
by Nigel Moriarty 02 Aug '13

02 Aug '13

Folks The latest edition of the Computational Crystallography Newsletter is available at http://phenix-online.org/newsletter The table of contents is listed below. Articles *cctbx* tools for transparent parallel job execution in Python. III. Remote access *phenix.ensemble_refinement*: a test study of apo and holo BACE1 Short Communications *CaBLAM*: Identification and scoring of disguised secondary structure at low resolution Structural Classification of Allergen IgE Epitopes by Hierarchical Clustering New Tool: *phenix.real_space_refine * 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

fix Ramachandran outliers
by Wei Shi 02 Aug '13

02 Aug '13

Hi all, I have a refined structure with phenix. The resolution=3.2 Å. After refinement, the R-work=0.2186, R-free=0.2864, Bonds=0.010, Angles=1.515. Ramachandran outliers: 4.8%, Ramachandran favored: 80.8%. Rotamer outliers: 14.5%, c-beta outliers: 2 clashscore: 16.28, overall score: 3.32 To fix the Ramanchandran outliers, in coot, by changing the phi-psi of the outliers, I could manage to move most of them to the allowed positions. But, after another refinement in phenix, those outliers come back. I am wondering whether any of you have any suggestions for me on how to fix these outliers. Also, I read in a previous post that “all outliers could be first fixed manually in Coot (also applying real-space refinement with Coot's Ramachandran restraints turned on), as the default potential is very tight and can pull residues the wrong way if they're starting from a very bad position”. I am wondering whether any of you have done this before in coot, and can you please tell me specifically how this is done. Thank you so much! Best, Wei

4 9

phenix-refine crash
by Carlos Frazao 01 Aug '13

01 Aug '13

Hi, After the last model building of my protein I am not succeeding to further refine its structure, because phenix.refine crashes. However, if I set "main.number_of_macro_cycles=0" the job reaches its expected end and outputs a .pdb file. If I set "main.number_of_macro_cycles=1" the job crashes a few lines after the log-file indication " ============================== Exporting results ==============================" and produces an incomplete .pdb file with part of the header. Alternatively, If I set "main.number_of_macro_cycles=2" the job crashes during "MACRO_CYCLE 2 OF 2" after log-file indication " ----------Correcting NCS rotamer outliers---------- Set U 69 GLU to tp10 rotamer" followed by the error message : "Traceback (most recent call last): File "/usr/local/phenix/phenix-dev-1391/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 "/usr/local/phenix/phenix-dev-1391/phenix/phenix/refinement/command_line.py", line 161, in run call_back_handler=call_back_handler) File "/usr/local/phenix/phenix-dev-1391/phenix/phenix/refinement/driver.py", line 1659, in run ncs_manager = self.ncs_manager).result File "/usr/local/phenix/phenix-dev-1391/phenix/phenix/refinement/strategies.py", line 170, in __init__ ion_placement_manager = ion_placement_manager) File "/usr/local/phenix/phenix-dev-1391/phenix/phenix/refinement/macro_cycle.py", line 66, in __init__ self.call(self.ncsrot, "NCS rotamer correction") File "/usr/local/phenix/phenix-dev-1391/phenix/phenix/refinement/macro_cycle.py", line 101, in call func(prefix = prefix) File "/usr/local/phenix/phenix-dev-1391/phenix/phenix/refinement/macro_cycle.py", line 146, in ncsrot log = self.log) File "/usr/local/phenix/phenix-dev-1391/phenix/phenix/refinement/misc.py", line 89, in ncs_rotamer_update log=log) File "/usr/local/phenix/phenix-dev-1391/cctbx_project/mmtbx/torsion_restraints/torsion_ncs.py", line 1111, in fix_rotamer_outliers key=key) File "/usr/local/phenix/phenix-dev-1391/cctbx_project/mmtbx/torsion_restraints/rotamer_search.py", line 178, in search assert rotamer == cur_rotamer AssertionError" I have been using built phenix-dev-1391 (I also checked phenix version 1.8.2-1309 or built dev-1436, they crashed as well) x86_64-fc3, in a machine running Ubuntu 10.04.4 LTS with Python 2.6.5. I tried another machine, running Ubuntu 12.04 LTS with Python 2.7.3, and using built phenix-dev-1391 x86_64-fc3 I got at the end of the log-file: " ----------Correcting NCS rotamer outliers---------- Traceback (most recent call last): File "/usr/local/phenix/phenix-dev-1391/build/intel-linux-2.6/../../phenix/phenix/command_line/refine.py", line 11, in <module> command_line.run(command_name="phenix.refine", args=sys.argv[1:]) File "/usr/local/phenix/phenix-dev-1391/phenix/phenix/refinement/command_line.py", line 161, in run call_back_handler=call_back_handler) File "/usr/local/phenix/phenix-dev-1391/phenix/phenix/refinement/driver.py", line 1659, in run ncs_manager = self.ncs_manager).result File "/usr/local/phenix/phenix-dev-1391/phenix/phenix/refinement/strategies.py", line 170, in __init__ ion_placement_manager = ion_placement_manager) File "/usr/local/phenix/phenix-dev-1391/phenix/phenix/refinement/macro_cycle.py", line 66, in __init__ self.call(self.ncsrot, "NCS rotamer correction") File "/usr/local/phenix/phenix-dev-1391/phenix/phenix/refinement/macro_cycle.py", line 101, in call func(prefix = prefix) File "/usr/local/phenix/phenix-dev-1391/phenix/phenix/refinement/macro_cycle.py", line 146, in ncsrot log = self.log) File "/usr/local/phenix/phenix-dev-1391/phenix/phenix/refinement/misc.py", line 89, in ncs_rotamer_update log=log) File "/usr/local/phenix/phenix-dev-1391/cctbx_project/mmtbx/torsion_restraints/torsion_ncs.py", line 1044, in fix_rotamer_outliers self.rotamer_search_manager.prepare_map(fmodel=fmodel) File "/usr/local/phenix/phenix-dev-1391/cctbx_project/mmtbx/torsion_restraints/rotamer_search.py", line 57, in prepare_map torsion_utils.prepare_map(fmodel=fmodel) File "/usr/local/phenix/phenix-dev-1391/cctbx_project/mmtbx/torsion_restraints/utils.py", line 771, in prepare_map not exclude_free_r_reflections)) File "/usr/local/phenix/phenix-dev-1391/cctbx_project/mmtbx/map_tools.py", line 242, in fft_map symmetry_flags = symmetry_flags) File "/usr/local/phenix/phenix-dev-1391/cctbx_project/cctbx/miller/__init__.py", line 3420, in fft_map f_000=f_000) File "/usr/local/phenix/phenix-dev-1391/cctbx_project/cctbx/miller/__init__.py", line 4236, in __init__ conjugate_flag=True) MemoryError"- I did not succeed to find out what may I have done wrong with the model and/or params files Any help is highly appreciated, Carlos -- ************************************** Dr. Carlos Frazao Structural Biology Laboratory - Macromolecular Crystallography Unit ITQB-UNL, Av Republica, Apartado 127 2781-901 Oeiras, P Phone: (351)-214469666 FAX: (351)-214433644 e-mail: frazao(a)itqb.unl.pt www.itqb.unl.pt

3 2

distance threshold for special positions?
by Alastair Fyfe 01 Aug '13

01 Aug '13

I'd be grateful for pointers to documentation/source regarding phenix.refine's criterion for flagging an atom's coordinates as coinciding with a special position. For shelxl I noticed a distance cutoff of 0.02 to the symmetry equivalent position seems adequate for the data I'm working with. For phenix.refine the threshold is higher but I haven't found documentation for the cutoff used. thanks, Alastair Fyfe

2 1

proteins sequence for phenix
by LISA 27 Jul '13

27 Jul '13

Dear All, Phenix will ask for the protein sequence when we create a new project. I try to solve my complex by the Autosol in phenix with proteins sequence. My complex contains 2 molecule A and 3 molecule B. How can I give the sequence in phenix? will it should be: >A .................................................................................................................. >B .............................................................................................................................. or >A .................................................................................................................. >A .................................................................................................................. >B .............................................................................................................................. >B .............................................................................................................................. >B .............................................................................................................................. Thank you. Lisa

3 2

Re: [phenixbb] On Mac two separate installations?
by Joseph Noel 27 Jul '13

27 Jul '13

Thanks Nat! I am trying out the new "Featured Enhanced Map" option. Preliminary analysis says it really helps! Also, following the protocol as outlined by Andy Karplus in his Science paper on data processing using data extending to CC1/2's near 0.2. Both together seem to really extend the resolution of my datasets and also improve refinement! ______________________________________________________________________________________ Joseph P. Noel, Ph.D. Arthur and Julie Woodrow Chair Investigator, Howard Hughes Medical Institute Professor, The Jack H. Skirball Center for Chemical Biology and Proteomics The Salk Institute for Biological Studies 10010 North Torrey Pines Road La Jolla, CA 92037 USA Phone: (858) 453-4100 extension 1442 Cell: (858) 349-4700 Fax: (858) 597-0855 E-mail: noel(a)salk.edu Publications & Citations: http://scholar.google.com/citations?user=xiL1lscAAAAJ Homepage Salk: http://www.salk.edu/faculty/noel.html Homepage HHMI: http://hhmi.org/research/investigators/noel.html ______________________________________________________________________________________

6 7

Fwd: Course on “Structure and Function of Membrane Proteins- a multidisciplinary approach”
by Nathaniel Echols 27 Jul '13

27 Jul '13

---------- Forwarded message ---------- From: Jose Artur Brito <jbrito(a)itqb.unl.pt> To: phenixbb(a)phenix-online.org Date: Fri, 26 Jul 2013 15:58:01 +0100 Subject: Course on “Structure and Function of Membrane Proteins- a multidisciplinary approach” Dear colleagues, We would like to announce the Instruct-sponsored 4-day course “Structure and Function of Membrane Proteins- a multidisciplinary approach”, to be held at ITQB, Oeiras (nearby Lisbon), Portugal, from 5 to 8 November 2013. The course will address several techniques for the structural and functional characterization of membrane proteins, with special emphasis on Macromolecular Crystallography, Mass Spectrometry, Electron Microscopy and Nuclear Magnetic Ressonance. The information provided by each technique, their applications in biological systems, major benefits and limitations will be discussed. The course is organized with lectures in the mornings and interactive tutorials in the afternoons. In addition, the selected applicants are grouped and each group works on a case study (based on 2-3 papers covering techniques of interest and a questionnaire) and gives a presentation at the end of the course. The course is aimed primarily at PhD students, young Post-Doctoral researchers and others with research activities involving structural biology of membrane proteins. Participants will be limited to 20. Candidate selection will be based on a short CV, a motivation letter from the applicant, and a recommendation letter from the supervisor. Registration deadline is September 16, please send your application to membstruct2013(a)itqb.unl.pt . More information can be found at http://www.itqb.unl.pt/membstruct2013 The organizing commission, Margarida Archer (ITQB-UNL), Carlos Cordeiro (FC-UL) and José A. Brito (ITQB-UNL) -- ************************************************ * José Artur Brito, PhD * * * * Post-Doctoral Fellow * * Membrane Protein Crystallography Lab * * Instituto de Tecnologia Química e Biológica * * Oeiras - Portugal * * * * Tel.: +351.21.446.97.61 * * Fax: +351.21.443.36.44 * * * * E-mail: jbrito(a)itqb.unl.pt * * URL: http://mx.itqb.unl.pt * ************************************************

1 0

Linux distribution question: who needs the FC3 build?
by Nathaniel Echols 26 Jul '13

26 Jul '13

Dear users, We are in the process of rearranging our server room and changing the way we do the nightly builds. One step of this will be retiring the bulky and ancient boxes which run Fedora 3, and standardizing on slightly more modern OSes. (Including Ubuntu-specific builds, as well as the usual Fedora/RedHat/CentOS base.) I know that some people have found that the Fedora 8 x86_64 build does *not* run properly on their systems, but the Fedora 3 build does. Could anyone who has firsthand experience with this please contact me off-list and tell me exactly which distribution and version you are running? We can't cover all possible combinations (sorry Gentoo users), but I'd like to get some idea what is still problematic. thanks, Nat

2 1

Refinement using a homolog as a reference
by Joshi, Amar (Dr.) 26 Jul '13

26 Jul '13

Hi, I am refining ligand-protein complexes at 3.2Å dataset with a human protein. My Rfree is ~0.34 but my ramachandran stats are concerning me: 85% favoured, 4.5% disallowed. The structure of a yeast homolog has been deposited at 2.4Å resolution. The yeast structure contains additional loops but is ~55% identical to the human protein. The overall conformation between the homologs is very similar. Can I use this as a reference model? Should I prune the reference model? Will different numbering schemes affect the refinement? Many thanks, Amar

4 3

Re: [phenixbb] Linux distribution question: who needs the FC3 build?
by Nathaniel Echols 25 Jul '13

25 Jul '13

On Wed, Jul 24, 2013 at 2:56 PM, Edward A. Berry <BerryE(a)upstate.edu> wrote: > Would that be the end of support for 32-bit linux? Or would you build > i686 version on a later version? (Or is it possible to run 64 bit > executable on 32- bit operating systems with some kind of compatibility > libraries and 64-bit capable hardware?) > > I have one Fc8 32 bit machine which I don't want to upgrade, but I don't > really need to run phenix on it. Just to clarify: we will continue distributing 32-bit Linux indefinitely (although probably on fewer platforms than 64-bit), and we will at least support Fedora 8 at a minimum for a while, since our most important development system is still running it. The question is really related to how ancient a distribution we need to support, and what scenarios are especially difficult for binary cross-compatibility. -Nat

1 0