September 2008 - phenixbb - phenix-online.org

memory error
by Anastasia A Aksyuk 22 Sep '08

22 Sep '08

Hello everyone! I have the following error when using phenix-1.3-final: /========================== Target weights: x-ray data ========================= Traceback (most recent call last): File "/usr/local/rossmann/phenix/phenix-1.3-final/phenix/phenix/command_line/refine.py", line 5, in <module> command_line.run(command_name="phenix.refine", args=sys.argv[1:]) File "/usr/local/rossmann.master/phenix/phenix-1.3-final/phenix/phenix/refinement/command_line.py", line 89, in run File "/usr/local/rossmann.master/phenix/phenix-1.3-final/phenix/phenix/refinement/driver.py", line 1081, in run File "/usr/local/rossmann.master/phenix/phenix-1.3-final/phenix/phenix/refinement/strategies.py", line 472, in refinement_machine File "/usr/local/rossmann.master/phenix/phenix-1.3-final/phenix/phenix/refinement/weight_xray_chem.py", line 264, in __init__ File "/usr/local/rossmann.master/phenix/phenix-1.3-final/phenix/phenix/refinement/weight_xray_chem.py", line 307, in xyz_weights File "/usr/local/rossmann.master/phenix/phenix-1.3-final/phenix/phenix/refinement/weight_xray_chem.py", line 196, in __init__ File "/usr/local/rossmann/phenix/phenix-1.3-final/mmtbx/mmtbx/__init__.py", line 72, in update_xray_structure force_update_f_mask = force_update_f_mask) File "/usr/local/rossmann/phenix/phenix-1.3-final/mmtbx/mmtbx/f_model.py", line 675, in update_xray_structure force_update = force_update_f_mask) File "/usr/local/rossmann/phenix/phenix-1.3-final/mmtbx/mmtbx/masks.py", line 207, in f_mask return self.compute_f_mask() File "/usr/local/rossmann/phenix/phenix-1.3-final/mmtbx/mmtbx/masks.py", line 222, in compute_f_mask miller_set = self.miller_array) File "/usr/local/rossmann/phenix/phenix-1.3-final/mmtbx/mmtbx/masks.py", line 138, in structure_factors flex.grid(fft_manager.m_real()).set_focus(fft_manager.n_real())) MemoryError /I am running with defaults, i.e. the following command: phenix.refine se.water2.noano_001-2acetate-noH-coot-0.pdb cr2remote.cr3.1.8-2_scaleit1.mtz ray_data.r_free_flags.test_flag_value=1 output.prefix=se.noano2.acetate refinement.input.xray_data.labels="F_peak(+),SIGF_peak(+),F_peak(-),SIGF_peak(-),merged" I would greatly appreciate your help to resolve this issue and I can provide any additional information needed. Thank you, Anastasia

2 3

phenix 1.3-rc4 vs. phenix 1.3-final
by Jianghai Zhu 19 Sep '08

19 Sep '08

Hi, I have done a couple of refinements using either phenix.refine 1.3-rc4 or phenix.refine 1.3-final. And phenix.refine 1.3-rc4 always gave me a lower Rfree, about 1% lower than phenix.refine 1.3-final, which actually is pretty big. What has changed between this two versions? -- Jianghai

4 4

Deborah Palestrant is out of the office.
by deborah.palestrant＠novartis.com 16 Sep '08

16 Sep '08

I will be out of the office starting 09/16/2008 and will not return until 09/18/2008. I will be checking email and will respond to your message at my earliest convenience.

1 0

Anomalous difference map file
by Andrzej Lyskowski 16 Sep '08

16 Sep '08

Hi, Is it possible to write an anomalous difference map file and not just the coefficients in the mtz file? Would it be possible to get a short instructions for the operations... Regards, Andrzej Lyskwoski -- Andrzej LYSKOWSKI, Ph.D. (andrzej[.]lyskowski[(a)]helsinki[.]fi) Institute of Biotechnology - Structural Biology & Biophysics P. O. Box 65 - Viikinkaari 1 - Floor/Room: 4/4316 Mail: FIN-00014 University of HELSINKI, FINLAND Delivery: FIN-00710 University of HELSINKI, FINLAND TEL.: 358 9 191 58921 FAX : 358 9 191 59940

3 2

phenix autobuild question
by Alex Berndt 16 Sep '08

16 Sep '08

Dear phenix community, I started an AutoBuild run using the phenix GUI (ver 1.3 rc4) on a linux cluster. It worked okay and all the refinement looked pretty decent. However, after quite a while I obtained the following error message (see below). Would anyone please tell me what to do to prevent this error and get phenix get its job finished? Also, I don't understand why phenix AutoBuild creates 5 (or whatever number) AutoBuild_run_x in the AutoBuild-run_1/TEMP0 directory and not two dirs up in the tree. In other words in my /someplace/home/someuser/ work/someproject/phenix2/AutoBuild_run_1_/TEMP0 directory are more AutoBuild_run_x_ directories (with x=1 to 5). It is a bit confusing to me. A final question: I realised that the phenix.refine drastically increases the number of outliers. I know that thare is a weighting term someplace ... but what was it again? Many thanks in advance, Alex warnings Failed to carry out AutoBuild_multiple_models: Sorry, subprocess failed...message is: ******************************************************************************** phenix.autobuild \ write_run_directory_to_file=/someplace/home/someuser/work/someproject/ phenix2/AutoBuild_run_1_/TEMP0/INFO_FILE_1 Reading effective parameters from /someplace/home/someuser/work/ someproject/phenix2/AutoBuild_run_1_/TEMP0/PARAMS_1.eff Sending output to AutoBuild_run_1_/AutoBuild_run_1_1.log ******************************************************************************** Failed to carry out AutoBuild_build_cycle: failure ******************************************************************************** ******************************************************************************** work_pdb_file "AutoBuild_run_1_/edited_pdb.pdb" working_directory "/someplace/home/someuser/work/someproject/phenix2" worst_percent_res_rebuild 2.0 --------------------------------------------------- Alex Berndt MRC-Laboratory of Molecular Biology Hills Road Cambridge, CB2 0QH U.K. phone: +44 (0)1223 402113 ---------------------------------------------------

5 5

BEAM TIME AVAILABLE FOR MACROMOLECULAR CRYSTALOGRAPHY
by Cyndi Salbego 15 Sep '08

15 Sep '08

BEAM TIME AVAILABLE FOR MACROMOLECULAR CRYSTALLOGRAPHY AT THE TWO ID BEAMLINES OF NE-CAT AT APS. Beamline 24ID-C: Variable energy between 6-18keV, ADSC Q315 Detector, MAD Capable. Beamline 24ID-E: Equipped with MD2 microdiffractometer, delivering beam as small as 5 microns. Single energy (12662eV), ADSC Q315 Detector. Contact csalbego(a)anl.gov for further details. See http://necat.chem.cornell.edu for information on our beamlines and a calendar of available days.

1 0

link in a double conformation residue
by mvliberato＠ursa.ifsc.usp.br 12 Sep '08

12 Sep '08

Dear all, I'm refining a protein structure with a cysteine covalently bonded with a fatty acid. However, this cysteine has a double conformation. I try to make a link with "cif_link.params", but I don't know how to describe the double conformation. How can I make this link between the fatty acid oxigen and the sulfur of the cysteine in one of the conformations? Any help would be much appreciated! Thanks Marcelo Marcelo Vizoná Liberato, Ms Physics Institute of São Carlos - USP São Carlos, Brazil ---------------------------------------------------------------- WEBMAIL - IFSC - Instituto de FÃsica de SÃ£o Carlos

2 2

Odd twinning issue (or maybe all twinning issues are odd by default)
by Francis E Reyes 11 Sep '08

11 Sep '08

All I had a structure that seemed perfectly twinned in P41 21 2, completely built to the observable density (no peaks in 2Fo-Fc map over 2sigma) and refined to 33/35 (R/Rfree). I indexed, scaled, and merged the original unmerged data to P41, molecular replaced using the above model. Three rounds of rigid body refinement of the MR solution with group adp's dropped R/Rfree to 28/30 (while generating new free R flags based on the P41 data). The resolution was to 2.5 so I skipped the NCS. Now I try to do a second round of refinement with this time individual_sites and individual_adps with no change to the input model. The final R/Rfree's are now in the 50's? Am I missing something (fundamental) ? Thanks FR --------------------------------------------- Francis Reyes M.Sc. 215 UCB University of Colorado at Boulder gpg --keyserver pgp.mit.edu --recv-keys 67BA8D5D 8AE2 F2F4 90F7 9640 28BC 686F 78FD 6669 67BA 8D5D

3 2

difficulties with autoMR
by Feldman III, George Franklin 10 Sep '08

10 Sep '08

I am trying to run phenix.automr on what I believe is perhaps a perfectly twinned data set, at 3A resolution. It is in space group 146 and according to the matthews has 3 molecules in the asu. This is a mutant of an orriginally published structure in which only one residue has been mutated (but crystallizes in a different space group). When I give the autoMR the following input: phenix.automr coords="MR.1.pdb" data=dtscale.ref resolution=0 component_type=protein identity="95" mass=13900 build=False copies=3 seq.dat I get an LLG of 37 and the run goes quite quickly. The speed of the run itself leads me to believe that I have a correct solution regardless of the slightly low LLG, but when I open the .pdb output file there is only one molecule in the asu. Am I doing something incorrectly? Why is there only one molecule instead of 3 in the asu, and how can I go about getting the other two to appear? I did let the program go ahead run autobuild, this resulted in a structure with an r_work:.5448 and r_free:.6005. I am relatively new to crystallography, but these values seem rather high to be a good starting point for refinement. I would appreciate any advice that might help me move on with my structure. Attached is the log file from phenix.xtriage in case it may be of any help. Thank you, in advance, for all of your help and advice. Best Regards, George F. Feldman

1 0

Re: [phenixbb] cci_apps on irix
by Ralf W. Grosse-Kunstleve 07 Sep '08

07 Sep '08

On 22 Aug 2008 o.h.weiergraeber(a)fz-juelich.de wrote: > I am just compiling the latest cci_apps package for the IRIX > platform. I know this is unsupported, but since the package is very > self-contained, I would not expect major problems with the command-line > packages (who needs a gui, anyway ;)) > I'm just running into two issues with the boost package which are > obviously a matter of syntax: Our IRIX machine is up again, and - to my surprise - the current CCI Apps compiled without errors. The IRIX binary bundle is available again at the usual place: http://www.phenix-online.org/download/cci_apps/ Ralf

1 0