April 2013 - phenixbb - phenix-online.org

Pause phenix runs
by Mark J van Raaij 13 Apr '13

13 Apr '13

Just wondering if being able to pause a phenix run would be something useful. Currently, it appears one can only abort. Currently doing refinement of a large structure, and doing several simultaneous runs with different options is slowing down the computer too much. Individual runs take about 72 hrs, so o/n is not an option (o/weekend will be...) (or is it possible and I have missed it?) if the FPLC can do it, why not phenix? Seriously, I don't know how easy or difficult it would be to implement, or if paused runs would still take up too many resources, but it seems like something that might be useful - i.e. set several long runs running, and temporarily interrupt them for other jobs on the same computer (like watching youtube movies, as some people seem to think I do all day...) Mark J van Raaij Lab 20B Dpto de Estructura de Macromoleculas Centro Nacional de Biotecnologia - CSIC c/Darwin 3 E-28049 Madrid, Spain tel. (+34) 91 585 4616 http://www.cnb.csic.es/~mjvanraaij

5 4

Problem with MR-Rosetta
by Marta Sanz Gaitero 12 Apr '13

12 Apr '13

Hi to everyone, I've have a little problem running MR-Rosetta in Phenix. Some weeks ago, I tried to solve some data which didn't have very much identity with any PDB model using MR-Rosseta. The program started running but after a few hours it stopped suddenly. To check if the problem was that it was not possible to solve my data because they were very complicated or if something was going wrong with MR-Rosetta, I decided to run it using model data. I runned it using the mtz of a protein we had already solved in the lab, its sequence, and its own PDB file, generated after solving it. After 10 minutes, the program stops when it's starting the rebuild step. I think the program shouldn't have any problem to solve this data, since the model used for molecular replacement is exactly the same protein in the mtz, so I wonder if i'm doing something wrong. If someone could gave me any idea which helped me to solve the problem i would be very grateful. I attatch a snapshot of the error. Thank you in advance, Regards, Marta

2 1

Re: [phenixbb] base stacking
by Ed. Pozharski 11 Apr '13

11 Apr '13

I wonder whether something simple like restraining a model to that of an ideal dna (a la den refinement) would do the trick with existing tools. -------- Original message -------- From: Nathaniel Echols <nechols(a)lbl.gov> Date: 04/10/2013 7:27 PM (GMT-05:00) To: PHENIX user mailing list <phenixbb(a)phenix-online.org> Subject: Re: [phenixbb] base stacking On Wed, Apr 10, 2013 at 4:09 PM, jp d <yoyoq(a)yahoo.com> wrote: is there a way to do base stacking in phenix ? If you mean restraints on base stacking, no, we don't have anything like this right now. The only implementation of this anywhere, as far as I know, is the one from your (Noller) lab, but (aside from the lack of an explicit license) the fact that it's derived from CNS makes me very reluctant to even look at the code. However, if anyone has the derivation for the restraint energy function, we'd be happy to consider adding this to Phenix - we need to make a lot of changes to how we handle base-pairing anyway. -Nat

2 1

Official release vs nightly builds
by Folmer Fredslund 11 Apr '13

11 Apr '13

Hi phenixbb/developers, Recently, some people have experienced problems with B-factor refinement in the official releases of PHENIX. It seems to me that, generally speaking, the first suggestion is to use the nightly builds of phenix. So, my question is: How does the official releases work? Is it a snapshot of the development version, or are you doing some kind of stabilization during the development of the official releases? Are you updating the official releases after release (to fix bugs) or are the bugs only fixed in the nightly builds. With the current official releases, would I be better off using the nightly builds for refinement? (I have also seen some weird behavior with B-factor refinement). Best regards and thank you for developing phenix, Folmer -- Folmer Fredslund Post doc MAX IV Laboratory Lund Sweden

6 11

Postdoc Position - Berkeley
by Paul Adams 11 Apr '13

11 Apr '13

Lawrence Berkeley Laboratory Postdoctoral Researcher Job ID: 75785 Division: Physical Biosciences Date Opened: 4/9/2013 Summary: A postdoctoral position is available to work on developing new methods for macromolecular structure determination and refinement with the PHENIX software system. To learn more about Phenix visit: http://www.phenix-online.org Duties: Development and implementation of computational algorithms for structure determination using crystallography. Including methods for the assessment of crystallographic data quality, improved methods for minimization of target functions, algorithms for the automated parameterization of structure refinement, new target functions for the refinement of structures, and enhanced structure deposition tools. Qualifications: Ph.D. in computational crystallography or a related field. Demonstrated experience with computational methods development. Publications in macromolecular crystallography or closely related field. Familiarity with software development tools such as C++, C and scripting languages such as Python. Strong programming and mathematical skills. Solid interpersonal skills and the ability to work in a team environment are critical. Ability to communicate with a broad range of researchers. Note that this is a multi-year position that is renewed annually on the basis on performance and continued funding. Lawrence Berkeley National Laboratory is a world leader in science and engineering research, with 13 Nobel Prize recipients over the past 75 years, and 57 present members of the National Academy of Sciences. LBNL conducts unclassified research across a wide range of scientific disciplines and hosts four national user facilities. AA/EEO employer committed to the development of a safe and diverse workforce. Learn more at http://www.lbl.gov. To apply visit https://lbl.taleo.net/careersection/2/jobsearch.ftl?lang=en and search for the job number 75785 or visit: https://lbl.taleo.net/careersection/2/jobdetail.ftl?lang=en&job=75785 -- 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 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

!!!
by Hexi Zhang 07 Apr '13

07 Apr '13

http://www.pispy.com/mmog/bupedamkfovmstocspifxj.bjyhn

1 0

Atoms with iso. B <=0
by Stephen Weeks 05 Apr '13

05 Apr '13

Dear phenixBBers I have a curious case of refinement of a 2.6 Angs structure, where some of the atoms have a calculated b-factor of 0.00. It seems to be somewhat random where this occurs and includes atoms within the backbone e.g. the nitrogen in the following residue: ATOM 2014 N THR A 264 -2.685 42.357 -35.833 1.00 0.00 N ATOM 2015 CA THR A 264 -3.952 41.776 -35.425 1.00 1.50 C ATOM 2016 CB THR A 264 -4.684 41.052 -36.593 1.00 1.70 C ATOM 2017 OG1 THR A 264 -4.077 39.772 -36.846 1.00 2.24 O ATOM 2018 CG2 THR A 264 -4.669 41.867 -37.855 1.00 1.72 C ATOM 2019 C THR A 264 -3.687 40.729 -34.368 1.00 2.42 C ATOM 2020 O THR A 264 -2.571 40.210 -34.263 1.00 2.63 O The structure contains 4 NCS related monomers but with a quick glance I don't see any obvious correlation between each one as to which atoms are 0.00. I did use phenix.pdbtools to randomise the B-factors (no atoms had a value 0.00 afterwards) but upon repeating the refinement the same problem occurs. Does anyone have a suggestion how to fix this ? Cheers Stephen

3 3

43rd Mid-Atlantic Macromolecular Crystallography Meeting
by Charles Pemble 05 Apr '13

05 Apr '13

*43**rd** Mid-Atlantic Macromolecular Crystallography Meeting* *May 30 – June 1, 2013* *Duke University* *Durham, NC* The 43rd Mid-Atlantic Macromolecular Crystallography Meeting (MAMCM) will be held at Duke University in Durham, North Carolina. This meeting fosters discussion and interaction among structural biologists from around the Mid-Atlantic region. Registration is now open at http://www.mid-atlantic.org. The deadline for registration is May 15th. This year will feature keynote speaker Tom Terwilliger from Los Alamos National Lab, workshops on PHENIX and MolProbity, plus many talks & posters on new and exciting research. Presentations will be selected from submitted abstracts and preference will be given to young researchers. The best presentation and poster by a Graduate Student & Postdoctoral Fellow will be awarded. We encourage you to submit your abstracts and sign up for the workshops as these are separate processes from registration. *IMPORTANT: *A block of rooms have been reserved at the Millennium hotel as well as the Duke dormitories. The deadline for reserving a room at the Millennium is May 9th. With best wishes on behalf of the primary Organizing Committee, Charlie Pemble _____________________________ Charles W. Pemble IV, Ph.D. Duke University Medical Center 308 Research Drive LSRC, A06 Durham, NC 27708 charles.pemble(a)duke.edu

1 0

table 1 no CC values, selenomethionine occupancy refinement
by Georg Mlynek 05 Apr '13

05 Apr '13

Dear All, I have 3 minor questions/problem and would be grateful for some help. 1a. Providing the XDS_ASCII.HKL to phenix tableI works perfect, however there are no values written for CC1/2. The same file in phenix.merging_statistics I get the CC1/2. Am I missing something? 1b. The only value phenix tablel one takes from INTEGRATE.LP (i guess CORRECT.LP will work too) is the wavelength - true? 2. I see some negative density around the seleniums in my selenomethionine labeled protein. I guess this is due to radiation damage. Is the accepted standard to do occupancy refinement for just the sidechain, or just parts of the sidechain, the whole residue or just leave it like it is. For the side chain refinement.refine.occupancies { constrained_group { selection = resname MSE ( name CB or name CG or name SE or name CE or name HB2 or name HB3 or name HG2 or name HG3 or name HE1 or name HE2 or name HE3 } } should work or? I am using version dev-1334 on ubuntu. Best regards Georg.

5 5

C-alpha r.m.s.d. calculation before and after structural alignment
by Bret Wallace 04 Apr '13

04 Apr '13

Dear All, I have solved a complex structure of two proteins, and a reviewer has asked to compare the altered interface binding of one particular binding partner (protein2) in this complex to various other similar deposited PDBs through r.m.s.d. calculations and rotation/translation matrices. After aligning ~4 structures from the PDB (specifically aligning 1 chain to the other binding partner, protein1) to my structure, I want to calculated the rmsd before and after re-aligning the deposited structures (containing protein2) to my complexed protein2. I have used the superpose utility in phenix to align the C-alphas, however, I am getting very large rmsd values after alignment (~3.3). This is rather strange as the aligned proteins are essentially the exact same, both in sequence and secondary structure. Using various different programs such as SSM, Coot, and others give me an rmsd value of ~1, which is far more reasonable, but they do not output a "start" rmsd value. Any suggestions on what the cause of this large rmsd value? Or possibly another program that could provide start and final rmsd values along with rotation/translation matrices for alignment? Thanks, Bret

2 1