- phenixbb - phenix-online.org

Re: [phenixbb] phenix.mr_rosetta not running "prerefine"
by Thomas C. Terwilliger 11 Oct '11

11 Oct '11

Hi Laurie, The problem is that mr_rosetta needed a pdb file name with more than 4 characters before the ".pdb" for this shortcut of generating pre-refined models to work (the Rosetta routine requires a longer file name). So if you rename your PDB file this should run (it does on my machine.) I will fix mr_rosetta so that it can take your PDB file. All the best, Tom T >> On Tue, Oct 11, 2011 at 12:33 PM, Thomas C. Terwilliger < >> terwilliger(a)lanl.gov> wrote: >> >>> Hi Laurie, >>> >>> Yes, it is not doing anything. There is something that didn't work in >>> the >>> inputs. Can you send me your fasta and PDB files and I'll just run it >>> here >>> to see what is wrong? >>> >>> Thanks! >>> -Tom T >>> >>> >> OK - I am also sending: >>> >> >>> >> >>> /netscr/bettsl/rosetta_test/MR_ROSETTA_2/GROUP_OF_PLACE_MODEL_1/RUN_1/REBUILD_IN_SETS_1/RUN_1/WORK_1/rebuild.log >>> >> >>> >> On Tue, Oct 11, 2011 at 12:01 PM, Thomas C. Terwilliger < >>> >> terwilliger(a)lanl.gov> wrote: >>> >> >>> >>> Hi Laurie, >>> >>> >>> >>> Yes, something didn't work it all right. Can you send me, off-list, >>> this >>> >>> log file: >>> >>> >>> >>> >>> >>> >>> /netscr/bettsl/rosetta_test/MR_ROSETTA_1/GROUP_OF_PLACE_MODEL_1/RUN_1/REBUILD_IN_SETS_1/RUN_FILE_1.log >>> >>> >>> >>> which I am guessing has some kind of error message in it. >>> >>> >>> >>> Thanks! >>> >>> -Tom T >>> >>> >>> >>> >> Hi - I tried the little prerefine without any data script in your >>> >>> >> documentation - I don't think it's working right. I attach the >>> >>> script >>> >>> >> file >>> >>> >> that I made (prerefine.dat) and the RUN_FILE_1.log. >>> >>> >> >>> >>> >> I can send more info offline if need be. >>> >>> >> >>> >>> >> Thanks >>> >>> >> >>> >>> >> LAurie Betts >>> >>> >> _______________________________________________ >>> >>> >> 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

refinements settings .eff
by Christian Roth 11 Oct '11

11 Oct '11

Hi Phenix gurus, I have done a refinement with an sodium ion. I had generated a phil file for the distance restraints to the surrounding water. Now i am not sure if it is really there. I removed it and removed the phil from the list in the gui. Now phenix complains about no Atom selected and stops. In the .eff file all restraints are incorporated and in I deleted therefore the geometric restraints manually from the . eff file. As a consequence an unused parameter warning appears about unused parameters or spelling errors in the input file. I guess the .eff file. My default parameter folder is empty. Does phenix load the .eff file from the previous run and would it help to remove this? I dont wanne start a new project just to be in the default configuration mode without additional information in the .eff file. I did this using the gui so I am not sure how to incorporate the --unused_ok command to which may help to convince phenix to process the job. I hope it is clear what I mean and thanks for any suggestion. Christian

2 1

phenix.mr_rosetta not running "prerefine"
by Laurie Betts 11 Oct '11

11 Oct '11

Hi - I tried the little prerefine without any data script in your documentation - I don't think it's working right. I attach the script file that I made (prerefine.dat) and the RUN_FILE_1.log. I can send more info offline if need be. Thanks LAurie Betts

2 1

Simulated annealing composite omit map conversion
by Yuri 11 Oct '11

11 Oct '11

The figures are good illustration of what seems to be the consensus: Finely sampled maps can represent the best mathematical approximation of the continuous electron density function. The weirdness I referred to probably has more to do with what Nathaniel alluded to previously. The fact that our eyes get so used to staring at coarse pointy density - caused by the default grid of building programs (in the case of "on the fly" map generation from coefficients). That of course can be changed if one changes the sampling rate and get their eyes used to the new look. that I suppose would become your new norm. -- Yuri Pompeu

1 0

MR_Rosetta and dimer of heterodimer
by Morten Grøftehauge 11 Oct '11

11 Oct '11

Hi everybody, I'm trying to phase a dimer of a heterodimer of a large, ~700aa, and small, ~150aa, component. The only good search model according to HHpred covers the first 500aa of the large component. So I am wondering whether I should add the 150aa to the end of the sequence file and add them to the fragment files in order for phenix.mr_rosetta to do density modification correctly? Or does this not matter at all? According to HHpred and others the last 200aa are beta sheets, probably a a beta sandwich or two, and the 150aa component consists of helical elements but that is as close as it really gets. Cheers, Morten

2 1

Re: [phenixbb] Simulated annealing composite omit map conversion
by Yuri 11 Oct '11

11 Oct '11

hi Nathaniel, Thats simple enough... Do I have any control over grid ( make a tighter looking mesh for publication) -- Yuri Pompeu

8 9

Phenix not reading my sca files
by Scott Classen 10 Oct '11

10 Oct '11

Why? Error is: Unknown file format for file /home/me/phenix/autosol/s4_peak.sca Unknown file format for file /home/me/phenix/autosol/s4_infl.sca Unknown file format for file /home/me/phenix/autosol/s4_rem.sca this also does not work: phenix.reflection_file_converter s4_peak.sca --mtz . error is: file_name: s4_peak.sca file_type: None No reflection data found in input file. Please advise. Thanks, Scott Classen

2 5

Phenix file organization
by R.M. Garavito 10 Oct '11

10 Oct '11

To the Phenix gurus, I am in the midst of reorganizing our little Mac computer cluster, and will have a mixed collection of machines with OS 10.6.8 or 10.7.1. While the reorganization is going fine for the most part, I am having a little trouble getting Phenix to work on one machine. The problem is integrating a machine (a MacPro) into the cluster with users having different UIDs from their NIS accounts on the rest of the cluster. Thus, the NIS UID leads to not only a different UID for the user, but a different account (and therefore a different "home" directory) name. Simply changing ownership of the user's directories and files globally in OS 10.6.8 leads to Phenix choking at some stage when used under the NIS UID, but without an informative error message. Things go on fine when used under the old local UID and user name. So my question is does Phenix store the absolute paths in its various project indices? Any suggestions would be helpful. Thanks, Michael **************************************************************** R. Michael Garavito, Ph.D. Professor of Biochemistry & Molecular Biology 513 Biochemistry Bldg. Michigan State University East Lansing, MI 48824-1319 Office: (517) 355-9724 Lab: (517) 353-9125 FAX: (517) 353-9334 Email: garavito(a)msu.edu ****************************************************************

2 1

Simulated annealing composite omit map conversion
by Yuri 10 Oct '11

10 Oct '11

"Fabricated" was probably a poorly chosen word. That implies made up data. I agree, I dont think cahnging the grid spacing would "introduce" unreal data to the maps... I just meant that it looks a little weird to have overly smooth mesh for a low res map... -- Yuri Pompeu

1 0

Computing Robs vs Rcalc for twinned data with xtriage
by Pietro Roversi 10 Oct '11

10 Oct '11

Dear all, I am trying to run phenix.xtriage to compute Robs vs Rcalc for twinned data (from A. A. Lebedev, A. A. Vagin and G. N. Murshudov Intensity statistics in twinned crystals with examples from the PDB, Acta Crystallographica, D62, 83-95, (2006) My input: phenix.xtriage xray_data.file_name=$1.mtz xray_data.obs_labels="IMEAN, SIGIMEAN" xray_data.calc_labels="FCalc, PHICalc" xray_data.high_resolution=3.1 > phenix.xtriage.$$.log The FCalc, PHICalc columns (from CCP4-sfall) are in my mtzfile: 1 ASC -27 27 0 100.00 0.7 10.4 52.72 2.60 H H 2 NONE -90 90 0 100.00 0.7 34.2 52.72 2.60 H K 3 NONE 0 15 0 100.00 5.7 5.7 52.72 2.60 H L 4 NONE 0.0 19.0 0 100.00 9.50 9.50 52.72 2.60 I FreeR_flag 5 NONE -244.3 13664.9 0 100.00 467.59 469.74 52.72 2.60 J IMEAN 6 NONE 1.8 990.9 0 100.00 43.05 43.05 52.72 2.60 Q SIGIMEAN 7 NONE 16.2 1157.9 0 100.00 173.14 173.14 52.72 2.60 F F 8 NONE 1.7 51.3 0 100.00 12.91 12.91 52.72 2.60 Q SIGF 9 NONE 0.7 7827.1 0 100.00 257.63 257.63 52.72 2.60 F FCalc 10 NONE 0.0 360.0 0 100.00 180.24 180.24 52.72 2.60 P PHICalc and yet I get the error message: creamint:~/Projects/FactorI/Tetartohedral/P1/DataFiles> ~/Scripts/phenix_xtriage.csh AUTOMATIC_DEFAULT_free.klh_sfall1 Running phenix.xtriage No matching array: scaling.input.xray_data.calc_labels=FCalc, PHICalc Possible choices: AUTOMATIC_DEFAULT_free.klh_sfall1.mtz:IMEAN,SIGIMEAN AUTOMATIC_DEFAULT_free.klh_sfall1.mtz:F,SIGF Please use scaling.input.xray_data.calc_labels to specify an unambiguous substring of the target label. Sorry: No matching array: scaling.input.xray_data.calc_labels=FCalc, PHICalc

2 1