August 2011 - phenixbb - phenix-online.org

unstable phenix?
by Kenneth A. Satyshur 26 Aug '11

26 Aug '11

We are running phenix (743) on a RHEL5 dual quad, multithread, making a composite omit map pf a very large structure (750 AA) and the system keeps getting locked up. Have anyone else seen this type of problem? Does multiprocessing work well on RHEL or just FC versions? We needed to install the FC8 version (/phenix-1.7.1-743/build/intel-linux-2.6-x86_64) to get it to work. So maybe we are using the wrong version? Shoiuld I recompile it for our system? Why is the gui always open when we are running long jobs (many days)? We need to put this stuff in the background and log out so someone else can run. Should scripts be used instead of the gui? Great program but too unstable. kas -- Kenneth A. Satyshur, M.S.,Ph.D. Associate Scientist University of Wisconsin Madison, Wisconsin 53706 608-215-5207

4 4

Weird Hydorgen refinement
by Yuri 24 Aug '11

24 Aug '11

Hi guys, I DEFINETELY ran refinement after I removed the hydrogens. My Rfree went up from 0.159 to about 0.167. Thanks for the tips. Ill give pdb.tools a try... -- Yuri Pompeu

3 2

Reference model restraints in autobuild?
by Francis E Reyes 24 Aug '11

24 Aug '11

Hi all While it is possible to generate refinement parameters to pass into autobuild, how does one pass the reference structure into the autobuild jobs if one were to refine using geometry restraints from the reference structure? (command line autobuild) If I can't supply the reference pdb, is there a phenix cmd utility that'll spit out all the restraints given a pdb so I can supply them in the refinement parameters file? F --------------------------------------------- Francis E. Reyes M.Sc. 215 UCB University of Colorado at Boulder

1 1

Re: [phenixbb] Weird Hydorgen refinement
by Yuri 23 Aug '11

23 Aug '11

Hi Pavel, The pdb did have anisou records for H's. I believe that is because in some previous runs I had only selected chain 'A' - I left out "and not elemnt H" - which I believe caused the H's to get refined anisotropically. I am trying to get rid of the ANISOU records and Ill run refinement again. ps. I am getting ready to deposit the coordinates and I dont want to keep hydrogens in the model, how can I get rid of them? I ve been using pyMol to remove them, but pymol gets rid of all the info -including symmetry -that phenix adds... Thanks -- Yuri Pompeu

3 2

Weird Hydorgen refinement
by Yuri 23 Aug '11

23 Aug '11

Hello everyone. I was trying to refine my model anisotropically. I selected element NA or (chain A and not element H). I looked at the output pdb shows that hydrogens were being refined anisou as well as the protein. I believe its because technically they are also in chain A, but shouldnt the default be never refine H's anisotropically, especially at 1.5A ?. -- Yuri Pompeu

2 1

autosol error in GUI
by r n 23 Aug '11

23 Aug '11

Hi all I downloaded the Phenix developing version 842, autosol gui running gave me the following error, "The key gui_output_dir is not in var_type_dict....command line wizzard needs conversion in the autosol.py script uing translating inputs or else in excluded inputs....wizard setVars routine needs this variable set". I have not had this error earlier version. any help? Thanks ram

2 1

pdb conversion to 3.2 format
by Luca Pellegrini 23 Aug '11

23 Aug '11

Hi, I am trying to use iotbx.pdb_remediator to convert PDB atom names to 3.2 format (as suggested by Nat Echols in a recent post to Coot mail-list). Unfortunately I get the following error message: lp212-mp2281 58% iotbx.pdb_remediator file_name=refine.pdb output_file=refine_3.2.pdb_file Traceback (most recent call last): File "/usr/local/cctbx/cctbx_sources/iotbx/command_line/pdb_remediator.py", line 54, in <module> run(sys.argv[1:]) File "/usr/local/cctbx/cctbx_sources/iotbx/command_line/pdb_remediator.py", line 51, in run remediator.remediator(params) File "/usr/local/cctbx/cctbx_sources/iotbx/pdb/remediation/remediator.py", line 257, in remediator user_dict) File "/usr/local/cctbx/cctbx_sources/iotbx/pdb/remediation/remediator.py", line 95, in build_hash test=os.path.isfile), "rb") TypeError: coercing to Unicode: need string or buffer, NoneType found It looks like a python problem, but I am afraid my insight ends here. Can anyone tell me how to fix this, please? I am not sure if it matters, but iotbx_pdb_remediator was not in my phenix installation (PHENIX-1.7.1-743 on MacOSX 10.6.8), so I installed cctbx separately. Thanks! Luca Luca Pellegrini Department of Biochemistry University of Cambridge 80 Tennis Court Road Cambridge CB2 1GA - UK Email: lp212(a)cam.ac.uk Tel: 0044-1223-760469 Fax: 0044-1223-766002 Sanger building, room 3.59

2 1

MR problem
by intekhab alam 22 Aug '11

22 Aug '11

Hi there I have a data of heterotrimer protein complex at 2.9A resolution. One protein consists of two domains. I tried phaser as well as molrep which gives a solution with only one domain. Rotation function and translation function were found to be fine for these solutions. I tried to find missing domain of the protein after fixing one of the domain or the other partner proteins using whole part or various truncations of missing domain. I also tried to find and build missing domain using Rosetta with the solutions of Molrep or Phaser as template. But, there were no solutions, or the solutions are clashed with other domain or proteins. Furthermore, R and Rfree is 30 and 40, respectively, and I could not reduce them further. There was almost no electron density map in the empty space so that I could not model manually. Plz guide me ,how can i look for the missing domain in the protein. regards -- INTEKHAB ALAM LABORATORY OF STRUCTURAL BIOINFORMATICS KOREA UNIVERSITY, SEOUL

2 1

matplotlib
by Ed Pozharski 21 Aug '11

21 Aug '11

This is really a rather esoteric issue, but I feel I need to mention this to save someone, depending on the situation, anywhere between five minutes of googling or two hours of hair-pulling. In a nutshell, phenix gui may in some circumstances screw up other programs that use matplotlib. I can't track down which specific phenix gui action causes it, but it's safe to assume that whenever matplotlib libs are used the subsequent calls to logarithmic scale plots (and likely other things) throw this error TypeError: coercing to Unicode: need string or buffer, dict found The solution is to erase the ~/.matplotlib/fontList.cache. The likely cause is that phinix gui calls on bundled matplotlib which is different from one I have installed (not to mention that I am using Lucid (because it's LTS) which has python 2.6 and not the 2.7 that is bundled with phenix). However, it still writes into the same ~/.matplotlib folder, thus I end up with incompatible data. Certainly, the problem will be gone when matplotlib gets bumped up to 1.0.1 in next Ubuntu release. This is yet another example of why the standalone installation approach is ideologically objectionable on modern Linux. But of course, the practical advantage gained by not having to package the software for any possible OS flavor/version users may choose outweighs the lower risks of package incompatibility and the reduced size of the packaged product. -- Oh, suddenly throwing a giraffe into a volcano to make water is crazy? Julian, King of Lemurs

5 5

formyl-methionine
by Yuri 20 Aug '11

20 Aug '11

Hello Everyone, Stupid question. I see density at my N-terminus that really looks like formyl-Met. I want to model it and refine. I notice REEl has FME ligand for the formyl-methionine, but I dont want to treat it as a "ligand" since it is covalently attached to SER 2. What is the best way to modify the MET residue? Cheers -- Yuri Pompeu

3 3