- phenixbb - phenix-online.org

Problem with mr_rosetta
by Kerff Fred 10 Sep '15

10 Sep '15

Hello, I’m trying to use mr_rosetta on my protein. I’m using OS X and upgraded phenix a couple of days ago and install rosetta at the same time. My job stops with the following message in this log file /phenix/MR_ROSETTA_52/GROUP_OF_ROSETTA_REBUILD_1/RUN_1/REBUILD_IN_SETS_1/RUN_6/WORK_1/rebuild.log. ERROR: can't find residue type at pos 1 in sequence ...sequence removed… ERROR:: Exit from: src/core/pose/annotated_sequence.cc line: 236 [35m0 mr_protocols.macosclangrelease 0x000000010b198bd6 print_backtrace() + 54 1 mr_protocols.macosclangrelease 0x000000010b198ab4 utility::exit(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, int, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, int) + 804 2 mr_protocols.macosclangrelease 0x000000010a9b8ca1 core::pose::residue_types_from_sequence(std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, core::chemical::ResidueTypeSet const&, bool) + 3697 3 mr_protocols.macosclangrelease 0x000000010a9ba6ea core::pose::make_pose_from_sequence(core::pose::Pose&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> > const&, core::chemical::ResidueTypeSet const&, bool) + 26 4 mr_protocols.macosclangrelease 0x0000000109267969 protocols::comparative_modeling::ThreadingJobInputter::pose_from_job(core::pose::Pose&, std::__1::shared_ptr<protocols::jd2::Job>) + 1449 5 mr_protocols.macosclangrelease 0x0000000109cb0c58 protocols::jd2::JobDistributor::run_one_job(std::__1::shared_ptr<protocols::moves::Mover>&, long, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >&, std::__1::basic_string<char, std::__1::char_traits<char>, std::__1::allocator<char> >&, unsigned long&, unsigned long&, bool) + 2232 6 mr_protocols.macosclangrelease 0x0000000109cafb5b protocols::jd2::JobDistributor::go_main(std::__1::shared_ptr<protocols::moves::Mover>) + 331 7 mr_protocols.macosclangrelease 0x0000000109c8d1f2 protocols::jd2::FileSystemJobDistributor::go(std::__1::shared_ptr<protocols::moves::Mover>) + 66 8 mr_protocols.macosclangrelease 0x00000001089e9964 my_main(void*) + 3124 9 mr_protocols.macosclangrelease 0x00000001089eae19 main + 4713 10 libdyld.dylib 0x00007fff8c8555c9 start + 1 11 ??? 0x0000000000000002 0x0 + 2 [0mError: ERROR: Exception caught by JobDistributor while trying to get pose from job 'S_PHENI_0001' Error: [ERROR] EXCN_utility_exit has been thrown from: src/core/pose/annotated_sequence.cc line: 236 ERROR: can't find residue type at pos 1 in sequence ...sequence removed... Error: Treating failure as bad input; canceling similar jobs protocols.jd2.FileSystemJobDistributor: job failed, reporting bad input; other jobs of same input will be canceled: S_PHENI_0001 protocols.jd2.JobDistributor: no more batches to process... protocols.jd2.JobDistributor: 15 jobs considered, 1 jobs attempted in 0 seconds caught exception 1 jobs failed; check output for error messages Anybody can help? Thanks Fred ----- Frédéric Kerff Chercheur qualifié F.R.S.-FNRS Cristallographie des protéines Centre d'Ingénierie des Protéines Université de Liège 17, Allée du 6 Août - Bat B5a 4000 Liège (Belgium) Tel.: +32 (0)4 3663620 Fax: +32 (0)4 3663772

1 0

refining different enantiomers
by Joel Tyndall 10 Sep '15

10 Sep '15

Hi all, I have a case where we have crystallised a ligand in our protein and the ligand purchased is a mixture of 4 structural isomers (enantiomers and diastereomers). There is no way of telling if one over the other is bound in the active site so we have assumed all 4 are binding. I have generated 4 separate ligands with 4 separate cifs and the all fit the density. I am refining the complex with all 4 ligands at 0.25 occupancy (occupancy refinement is switched off, I have changed the clash guard non-bonded distance threshold is set to 0.0 (as an initial error message came up with nonbonded interactions < 0.5). My refinement ran to completion, but something is definitely not right. The pdb file won't load in Coot and in pymol the ligands have imploded/exploded. I am just wondering if this is the best way to refine this structure (or I have missed something) and probably more to the point what have I missed with the ligands. To me it is almost identical to alternate conformations, but I have obviously missed something I have just install phenix 1.10 on a windows machine. Thanks J _________________________________ Joel Tyndall, PhD Associate Professor in Medicinal Chemistry National School of Pharmacy University of Otago PO Box 56 Dunedin 9054 New Zealand Skype: jtyndall Ph: +64 3 479 7293

1 0

Phenix version 1.10 released
by Paul Adams 09 Sep '15

09 Sep '15

The Phenix developers are pleased to announce that version 1.10 of Phenix is now available (build 1.10-2155). Binary installers for Linux, Mac OSX, and Windows platforms are available at the download site: http://phenix-online.org/download/ Highlights for this version: CDL as the default protein restraints, tools for planning a SAD experiment and scaling/merging SAD data, refactored xtriage GUI, use of intensities in molecular replacement, occupancy refinement mode after MR to automatically prune the model, expanded real space refinement for cryo-EM (and X-ray) data, NCS constraints in refinement, improved rotamer fitting in refinement, improved and consolidated secondary structure restraints, validation of peptide planes, protein backbone validation using Cablam, structure search tool, updated feature enhanced map protocol, TLS analysis tools, new continuous integration build system (BuildBot), and many other new utilities and fixes. General: ======== - New continuous build system, Buildbot - New tools for planning and evaluating SAD data collection and SAD datasets - All command-line installers converted to Python - Removed dependency on Apple tools for Mac installer relocation - Improved geometry restraints: - Nucleic acids: basepair planarity, hydrogen bonding and stacking restraints - Angle and torsion restraints for disulfide bonds - Conformation-Dependent Library is now the default (cdl=True) - Automatic linking has been expanded with carbohydrate linking being the default in all applicable programs - Names and places of parameters controlling secondary structure, reference and NCS restraints were changed - NCS constraints ("strict" NCS) for refinement (phenix.refine and phenix.real_space_refine) - Applied to coordinates and ADPs - Option to refine NCS operators - Partial LINK records support: - only in phenix.refine, phenix.geometry_minimization & phenix.real_space_refine - output LINK records generated using automatic linking and applied .edits (LINK in input PDB file are ignored) - HELIX/SHEET records in input file will be preserved in output PDB file Graphical interface: ==================== - Integrated help browser now supported on Linux - Improved display of Xtriage results, with summary of problems with data - Support for 'within' syntax in graphical atom selections - Model Reconstruction GUI: new GUI for building complete asymmetric unit and biological molecules using MTRIX and BIOMT records, respectively, in the PDB file - New interface for phenix.anomalous_signal - New interface for phenix.scale_and_merge Refinement: =========== - phenix.refine - Optionally use AFITT for geometry restraints of ligands - mmCIF output now includes _refln.pdbx_r_free_flag tag to explicitly record original R-free flags (instead of implicitly in status tag) - Updated Phenix GUI to handle reorganization of NCS implementation - main.secondary_structure_restraints, main.ncs and main.reference_model_restraints parameters were removed - Improved rotamer fitting (faster, less outliers) - Bug fixes - phenix.real_space_refine: - rigid-body refinement with possibility to specify rigid groups - NCS constraints are used by default - NCS groups are found automatically, also can be specified manually - support for custom restraints - support for ligand CIF files - map can have any origin (non-zero), refined model is not shifted to zero origin - if CCP4 map is used resolution must be specified using “resolution" keyword - Enable reference model restraints, including restraining to self - Many bug fixes, many tests added - phenix.secondary_structure_restraints (proteins and DNA/RNA) - Three methods are available to annotate protein secondary structure - Protein secondary structure definition may be outputted as HELIX/SHEET records in PDB format - New algorithm to search for nucleic acid base pairs and stacked pairs. - phenix.rosetta_refine: - more options for map type for real-space refinement, including composit omit and feature-enhanced maps - partial support for ligands (requires separate parameter files) - phenix.tls_analysis: - Decomposition of TLS matrices into elemental rigid-body motions (axes and amplitudes of vibration and libration). Validate TLS matrices. - phenix.tls_as_xyz: - Explicit interpretation of refined TLS parameters by decomposing into elemental rigid-body motions and generating structural ensembles that are consistent with these motions - phenix.multistart_sa: - utility for running multi-start simulated annealing refinements with varying random seed - sample uncertainty due to stochastic effects - also performs map averaging - parallelizes across multicore systems or managed clusters Experimental Phasing and Model Building: ======================================== - phenix.xtriage: - refactored for greater modularity and consistency between command line and Phenix GUI - simplified and clarified feedback, including new summary panel - many bug and stability fixes from PDB-wide testing - detection of elliptical truncation - analyze completeness of anomalous vs non-anomalous dataset - phenix.plan_sad_experiment (new command, GUI and command-line): - Estimate the I/sigI for a dataset that will be necessary to solve a structure by SAD - phenix.scale_and_merge (new command, GUI and command-line): - scale one or more SAD datasets, optimally using unmerged original index data, to create a scaled, merged dataset and to estimate the half-dataset anomalous correlation - phenix.get_cc_ano (new command,command-line only): - Calculate the anomalous correlation between two datasets - phenix.get_patterson_skew (new command, command-line only): - Get the skew of an anomalous difference Patterson and the ratio of anomalous differences to anomalous error. - phenix.merging_statistics: - calculate CC(anom) for split half-datasets Molecular Replacement: ====================== - Phaser: - now allows direct use of intensities instead of amplitudes for molecular replacement - new treatment of effect of intensity measurement error introduced - improved handling of weak data in normalization and anisotropy correction - new occupancy refinement mode to retain or prune residues along a chain - highlights loops and domains that are likely to need major correction - automatically invoked if high TFZ-score solution rejected for bad packing - subgroups feature: Phaser lists possible subgroups of space group, and if twinning is suspected the space group can be set explicitly to one of its subgroups for a molecular replacement search. In this case, the data will be expanded to the unique set for the subgroup. - new "Search" panel GUI making it less confusing to define multicomponent searches - replacing "Composition" with "ASU contents" on the GUI - phased translation function can be used, even as part of MR_AUTO search - magnification factor (cell scale) can be refined for electron density as a model: useful when cryoEM image reconstructions serve as models - SAD likelihood now reported as log-likelihood-gain: positive number relative to null hypothesis - Phaser now prints out suggestion for number of copies related by successive application of NCS translations (TNCS NMOL parameter), if suggested value > 2 - fixed bug affecting xyz-values > 100 in PDB for substructure - various other bug fixes - phenix.find_alt_orig_sym_mate: - Created new alias, phenix.famos, which is an anagram of the acronym for find_alt_orig_sym_mate and is easier to remember for the commandline. - phenix.subgroup_symmetry_datasets: - Calculates all subgroup of the diffraction data, and outputs them on the conventional setting. Useful to explore options when multple twin laws are possible. - phenix.structural_domain_search: - Domain search program using an approach based on graph theory. Very quick and fairly accurate for simple problems. Validation: =========== - phenix.omegalyze: - Identify non-trans peptide bonds in proteins. Three categories of peptide bond geometry: trans, cis, and twisted (>30deg away from planar). - phenix.cablam_validate: - Validation of protein backbones now uses separate contours for trans-proline vs cis-proline. - phenix.molprobity: - incorporated RNA validation New commands: ============= - phenix.model_idealization (new command, command-line only): - phenix.plan_sad_experiment (new command, GUI and command-line): - phenix.scale_and_merge (new command, GUI and command-line): - phenix.get_cc_ano (new command,command-line only): - phenix.get_patterson_skew (new command, command-line only): Other: ====== - phenix.model_idealization (new command, command-line only): - Idealize geometry of the input model. Only secondary structure elements of proteins are currently supported. - phenix.structure_search Quickly find homologous structures for a given model. There is an option to compile and output a list of ligands found in structures of those homologs. - phenix.map_model_cc (and its synonym phenix.model_map_cc): Compute model-map correlation coefficient (map CC or RSCC). Output overall, per whole model, per chain and per residue CC. Input map needs to be CCP4 fromatted file. - phenix.fem: - various bug fixes - make using "maximal synthesis" the default (before "minimal synthesis" was the default) - current version obsoletes all previous versions - phenix.chains_as_models: Convert PDB file with multiple chains into PDB file with multiple models (MODEL-ENDMDL). - phenix.models_as_chains: Convert multi-model (MODEL-ENDMDL) file into one model file with each model assigned unique chain ID. - phenix.helix_sheet_recs_as_pdb_files: Split PDB file into several PDB files with each file representing HELIX or SHEET record found in PDB file header. - phenix.map_value_at_point: Output map values at specified (x,y,z) points in space. - phenix.map_box: Extract box around selected atoms with map. Output PDB with atoms in the box and map files in CCP4, X-plor and MTZ formats. - phenix.model_model_distances: Given two PDB files output distances per atom, residue, chain, model and overall. It is assumed (and asserted in the code!) that the amount and order of atoms in both files are identical. - phenix.simple_ncs_from_pdb: Tool to find NCS groups in PDB model. Available in previous versions. Major code re-write to improve re-usabiliy. For a full list of changes see: http://www.phenix-online.org/documentation/CHANGES Please note that this publication should be used to cite use of Phenix: PHENIX: a comprehensive Python-based system for macromolecular structure solution. P. D. Adams, P. V. Afonine, G. Bunkóczi, V. B. Chen, I. W. Davis, N. Echols, J. J. Headd, L.-W. Hung, G. J. Kapral, R. W. Grosse-Kunstleve, A. J. McCoy, N. W. Moriarty, R. Oeffner, R. J. Read, D. C. Richardson, J. S. Richardson, T. C. Terwilliger and P. H. Zwart. Acta Cryst. D66, 213-221 (2010). Full documentation is available here: http://www.phenix-online.org/documentation/ There is a Phenix bulletin board: http://www.phenix-online.org/mailman/listinfo/phenixbb/ Please consult the installer README file or online documentation for installation instructions. Direct questions and problem reports to the bulletin board or: help(a)phenix-online.org and bugs(a)phenix-online.org Commercial users interested in obtaining access to Phenix should visit the Phenix website for information about the Phenix Industrial Consortium. The development of Phenix is principally funded by the National Institute of General Medical Sciences (NIH) under grant P01-GM063210. We also acknowledge the generous support of the members of the Phenix Industrial Consortium. -- 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 33, Room 347 Building 80, Room 247 Building 978, Room 4126 Tel: 1-510-486-4225, Fax: 1-510-486-5909 http://cci.lbl.gov/paul Lawrence Berkeley Laboratory 1 Cyclotron Road BLDG 33R0345 Berkeley, CA 94720, USA. Executive Assistant: Louise Benvenue [ LBenvenue(a)lbl.gov ][ 1-510-495-2506 ] --

4 3

on "occupancies" in the phenix.refine GUI
by Smith Liu 08 Sep '15

08 Sep '15

Dear All, In my map for a specific HISresidue, it was rather clear the sidechain has 2 conformations. In my start modle, I only have the PDB for 1 of the 2 conformations of the sidechain. With "occupancies" clicked, the refined PDB still only contain the PDB for 1of the 2 conformations of the sidechain (occupancy is 1). Thus in order to have the GUI "occupancies" takes function, I need manually add the PDB for the conformation of the other side chain, am I right? If I am right，will you please show me an PDB example which contains the conformations for 2 or several sidechains of a residue, so that I can know how to edit this kind of PDB? Best regards. Smith

2 1

Transfer FreeR and FreeR test set flag value
by Matthew Chu 07 Sep '15

07 Sep '15

Dear All, I am trying to refine my model against a higher resolution dataset. My colleage's old-low resolution dataset was processed in HKL/scalepack and the FreeR was assigned by Phenix.reflection file editor, which I assume the test set was marked as "1". Now I got a new-higher resolution dataset, and plan to process it in XDS/Aimless. I want to transfer the FreeR set from his old mtz and also use his low res model for the new refinement. My question is, when I transfer the FreeR set from his old mtz with flag value as "1" in Aimless, is the flag value now "1" or "0" in the new mtz output by Aimless? Can Aimless recognize the FreeR flag from the old mtz which are marked as "1"? I also read an old post from Nat Echols in 2011: "For Phenix, 1 is test, 0 is work, but you could reverse this and it will still figure out what to do automatically", so does it mean it doesn't matter if the FreeR is "1" or "0" in the mtz from Aimless, if I use phenix.refine for my new dataset, Phenix can figure out which is the correct FreeR value? It's a mix of CCP4 and Phenix question, so I have crossed posted it in the CCP4bb as well. Many thanks, Matt

2 2

a question on phenix refine
by Smith Liu 06 Sep '15

06 Sep '15

Dear All, For the GUI of phenix.refine, if we click "Optimize X-ray/stereochemistry weight", it should optimize on how to part the weight between X-ray and stereochemistry. Then if we do not click "Optimize X-ray/stereochemistry weight", phenix.refine will do with whole weight from "X-ray" or whole weight from "stereochemistry"? I am looking forward to getting your reply. Smith

2 1

Space group
by mohamed noor 04 Sep '15

04 Sep '15

Dear all I processed my images in point group P 3 2 1. After running Phaser and building it, the space group was P 3(2) 2 1. However, the refinement is stuck with Rfree at around 35-38 % (data extending to 3.3 A). As I believe the R factor is a bit high, and I have a different crystal form in P4(2) 2(1) 2 with R free at 25 % for data to 3.8 A, I expanded it to P 1, redid Phaser searching for 48 copies. After refinement, the R free is at 38 %. For some reason, Zanuda/Refmac is not able to read my files. So I ran Xtriage instead: A summary of R values for various possible point groups follow. ------------------------------------------------------------------------------------------------------- | Point group | mean R_used | max R_used | mean R_unused | min R_unused | BIC | choice | ------------------------------------------------------------------------------------------------------- | P 3 | 0.000 | 0.000 | 0.271 | 0.000 | 3.521e+05 | | | C 1 2 1 (2*y,-x+y,z) | 0.423 | 0.423 | 0.200 | 0.000 | 1.763e+06 | | | P 1 | None | None | 0.219 | 0.000 | 1.056e+06 | | | P 6 | 0.242 | 0.423 | 0.200 | 0.000 | 1.365e+06 | | | C 2 2 2 (x+y,2*y,z) | 0.271 | 0.423 | 0.200 | 0.000 | 1.464e+06 | | | P 3 2 1 | 0.000 | 0.000 | 0.423 | 0.423 | 1.761e+05 | | | P 3 1 2 | 0.242 | 0.423 | 0.200 | 0.000 | 1.365e+06 | | | C 2 2 2 (2*x,x+y,z) | 0.271 | 0.423 | 0.200 | 0.000 | 1.464e+06 | | | P 1 1 2 | 0.423 | 0.423 | 0.200 | 0.000 | 1.763e+06 | | | P 6 2 2 | 0.219 | 0.423 | None | None | 1.265e+06 | | | C 1 2 1 (x+y,-x+y,z) | 0.000 | 0.000 | 0.242 | 0.000 | 5.282e+05 | | | C 1 2 1 (x+y,2*y,z) | 0.423 | 0.423 | 0.200 | 0.000 | 1.763e+06 | | | C 1 2 1 (x-y,2*x,z) | 0.000 | 0.000 | 0.242 | 0.000 | 5.282e+05 | | | C 2 2 2 (x-y,x+y,z) | 0.271 | 0.423 | 0.200 | 0.000 | 1.464e+06 | | | C 1 2 1 (2*x,x+y,z) | 0.000 | 0.000 | 0.242 | 0.000 | 5.282e+05 | | | C 1 2 1 (x-y,x+y,z) | 0.423 | 0.423 | 0.200 | 0.000 | 1.763e+06 | | ------------------------------------------------------------------------------------------------------- R_used: mean and maximum R value for symmetry operators *used* in this point group R_unused: mean and minimum R value for symmetry operators *not used* in this point group Looking at the R values, it is possible that it is actually C 1 2 1. My questions: 1. What is the difference between the C 1 2 1 lines? What do the values in brackets mean? 2. Since Xtriage did indicate a possible NCS axis parallel to crystallographic axis (and it looks almost like that in my P321 model), do I reintegrate in one of those C121 and then redo Phaser/refinement? I can post a screenshot of model if that will help. Thanks. Mohamed

1 0

Positive density at metal ions
by Jan van Agthoven 04 Sep '15

04 Sep '15

Hi everyone, I work on a structure of about 3 A resolution. The structure is almost done except for 3 metal ions which remain positive in the density. The metal is clearly Mn (refinement works fine for an other very similar structure) and has been set at full occupancy. I tried to reset the B factor at lower values, which solves the problem for 2 metal ions but not the 3rd one. Is there anything else I can do? Jan

2 1

Rwork higher than Rfree
by Smith Liu 01 Sep '15

01 Sep '15

Dear All, Is poor resolution a factor leading to Rwork higher than Rfree? Any other factor? Best regards. Smith

1 0

Cell dimentions
by Bradley Hintze 30 Aug '15

30 Aug '15

Hello I have some interesting DNA structures I would like help with . The duplexes are stacked in the crystal head to head and tail to tail. This creates an interesting situation where an adenine is base pairing with itself across cell symmetry. The most likely way in which adenines base pairs with itself is if one is in the syn conformation and the other is in anti, i.e. a Hoogsteen base pair. However, that would break the symmetry as they are the same adenine and thus can't have two conformations. As such, I think it's appropriate to have two molecules, rather than one, in the asymmetric unit. I can make the new model easily enough by generating a symmetry mate and assigning different chain IDs but how do I appropriately calculate the new cell dimensions? Thanks, Bradley

2 1