phenix_logo
Python-based Hierarchical ENvironment for Integrated Xtallography
Documentation Home
 

The phenix graphical interface

Overview
Running programs from the GUI
Modules
Reflection tools
Model tools
Phasing
Molecular replacement
Model building
Refinement
Maps
Ligands
Validation
Utilities
Other tools
Projects
Running PHENIX tutorials
Preferences
Coot integration
PyMOL integration
Reporting bugs
Acknowledgments

Overview

The Phenix GUI is primarily a frontend to the command-line programs, with several extra graphical utilities for validation, map generation, and file manipulations. The original GUI is still available but deprecated, and all functionality has been more or less superseded. This page covers the main interface and common behavior; individual program GUIs are covered separately.

The main GUI is started simply by typing the command phenix; by default, it will also open automatically when you launch any of the individual program GUIs. You should not run more than one instance of the GUI at a time, to avoid conflicts with internal database files.

images/gui_main.png

Running programs from the GUI

When starting a job, Phenix writes out a configuration file and calls the command-line version of the program. By default, this is started directly in the main process, i.e. "locally", which allows communication between the program and the GUI in memory rather than via temporary files. The drawback to this is that if the GUI is closed or crashes, the job will be ended too. An alternate "detached" mode is available, which starts the job as an entirely separate process. This limits the speed at which the GUI can be updated, but allows quitting the GUI without stopping the job.

A third mode is to run jobs on a queueing system; while this could be done entirely on a multi-core workstation, it will usually be spread across a cluster of similar computers. Currently only Sun Grid Engine is supported. To enable queued jobs, open the Preferences, switch to the "Processes" tab, and check the box to enable queueing. The queue submission option will now appear when starting a job. For this to work, the SGE binaries ('qsub', 'qstat', and 'qdel') must be in the current environment $PATH, Phenix must be installed in the same location on all nodes, and the filesystem on which the job is being run must be mounted on all nodes. The GUI will display a "waiting" status until the job is actually started. The main interface has a window for viewing the current queue status (Utilities->Show queue status).

Modules

Individual programs are grouped by category. Except where noted, most of these correspond to command-line programs, and the documentation for the command-line version should be the primary reference for understanding program behavior and inputs. Additional GUI documentation is available for some programs.

Reflection tools

  • Xtriage: reflection data analysis and quality assessment; used to detect twinning and other pathologies.
  • Reflection file editor: utility for merging reflection files and creating or extending R-free flags. (GUI only)
  • Calculate F(model): utility for generating structure factors (as real or complex numbers) from a model alone.
  • Import CIF structure factors: convert a CIF file (used by the PDB to store deposited experimental data) to an MTZ file. This will automatically fetch data for a known ID from the PDB, or you may supply your own files.
  • French & Wilson data correction: procedure for estimating appropriate values for weak and negative intensities. This is also run automatically as part of several programs, including phenix.refine and the Phaser GUIs.
  • Model-based phases: calculate phases and structure factors in a variety of formats (including Hendrickson-Lattman coefficients).
  • 3D and 2D data viewers: these display data from reflection files (usually amplitudes or intensities) as they appear in reciprocal space, either in a full 3D view or a pseudo-precession camera section.

Model tools

  • PDB Tools: various model manipulations such as modifying atom records, geometry minimization, and generating "fake" structure factors.
  • Combine PDB files: simple utility for merging models split over several PDB files (such as massive complexes like the ribosome). Chain IDs will be modified if necessary (using up to two characters).
  • Superpose PDB files: structure alignment program.
  • NCS identification: determine appropriate NCS operators and/or restraint groups based on a set of heavy-atom sites, macromolecule chains, or electron density.
  • Apply NCS: apply NCS operators to a PDB file to generate the complete asymmetric unit.

Note that a simpler NCS GUI is also available as a plugin to the phenix.refine GUI, and does not necessarily need to be run separately.

Phasing

  • AutoSol: automated experimental phasing for all experiment types (SAD, MAD, MIR, etc.); also performs simple model-building after phasing. Combines HySS, Phaser, SOLVE, RESOLVE, and phenix.refine.
  • Hybrid substructure search: heavy-atom site identification program. Used automatically as part of AutoSol.
  • Phaser-EP: Another Phaser interface, for SAD and MR-SAD phasing. We recommend trying AutoSol first, but this GUI exposes additional parameters.

Molecular replacement

  • AutoMR: automated molecular replacement using Phaser; feeds directly into AutoBuild. (This is also the starting point for running MR-SAD.) There are two interfaces, one for simple structures with one or more copies of a single chain, and a more complex interface for heterogeneous structures.
  • Phaser-MR: Interface for molecular replacement using standalone Phaser, with all parameters available as well as different modes of operation. We recommend starting with AutoMR first, but this GUI is useful for tough cases.
  • MR-Rosetta: automation pipeline for exceptionally difficult structures, which uses the Rosetta software for protein structure prediction and design to rebuild poor MR solutions, along with Phaser and AutoBuild. (Separate installation of Rosetta is required.)
  • Sculptor: prepare a search model for molecular replacement by trimming the structure, modifying B-factors, etc.
  • Sculptor - Coot interface: a Coot plugin for running Sculptor interactively and visualizing results.
  • Ensembler: tool for creating superimposed ensembles of related MR search models.

Model building

  • AutoBuild: automated model-building and refinement with RESOLVE and phenix.refine. This GUI only encapsulates the building functions; a separate GUI is available for omit-map calculations.
  • Phase and build: similar to AutoBuild, but significantly faster (although less accurate).
  • Find Helices and Strands: fast building of secondary structure (as poly-ALA) into maps with RESOLVE and PULCHRA. Especially good for low-resolution data.
  • Fit Loops: simple GUI for building missing loops into a map based on a sequence and start model.
  • Morph model: improvement of poor molecular replacement solutions by "morphing" into density and rebuilding.

Refinement

  • phenix.refine: automated refinement, supporting both X-ray and neutron data. In addition to the features available in the command-line program, the GUI version includes graphical atom selection, simplified setup of restraints, automatic addition of hydrogens, and post-refinement validation.
  • ReadySet: preparation of input files for refinement, including generating restraints (CIF) files and adding hydrogens.

Maps

  • Create map coefficients: simple GUI for generating likelihood-weighted maps, including 2mFo-DFc, mFo-DFc, anomalous, and others, including averaged "kick" maps. Currently works only on untwinned data. (GUI only)
  • FFT map coefficients: convert an MTZ file containing map coefficients to one or more CCP4 or XPLOR-format maps suitable for viewing in PyMOL, etc. (Note that the main maps GUI and many other programs can also output these files, and the GUI will generate them for you as necessary for PyMOL.)
  • AutoBuild - create omit map: similar to the main AutoBuild GUI, but used for omit map generation rather than building.
  • RESOLVE density modification: This is just a simplified interface for running the AutoBuild wizard in map-only mode.
  • Isomorphous difference map: simple utility for creating a map from two sets of amplitudes. Corresponds to the command-line program phenix.fobs_minus_fobs_map.
  • Superpose maps: given two PDB files and accompanying map coefficients (in MTZ format), superposes the PDB files and reorients the maps to follow them. Output is the reoriented PDB files and maps in CCP4 format.
  • Cut out density: extracts a section of electron density from input map coefficients, and outputs new map coefficients (suitable for running molecular replacement in Phaser).
  • Multi-crystal averaging: performs density modification on different crystal forms of the same structure and averages the maps around the models.

Ligands

  • LigandFit: wizard for placing ligands in electron density maps, accounting for necessary conformational changes.
  • eLBOW: electronic Ligand Builder and Optimization Workbench, a tool for generating restraints (and geometries) for any molecule, using a variety of inputs including PDB file and SMILES string.
  • REEL: "Restraints Editor Especially Ligands", a graphical editor for CIF files, also serves as a frontend to eLBOW. Currently runs as a separate program. (GUI only)
  • Ligand search: program for identifying unknown blobs of electron density and placing appropriate ligand(s).

Validation

Although these programs exist only in the GUI, some of the individual analyses are available as command-line tools (phenix.ramalyze, phenix.rotalyze, phenix.cbetadev, and phenix.clashscore). Because all three programs are subsets of the same functionality, their documentation is lumped together.

  • Comprehensive validation: reports R-work and R-free, statistics for geometry restraints, Ramachandran plot, sidechain rotamers, C-beta deviation, all-atom contacts, and real-space correlation with electron density. Outlier lists are linked to graphics programs such as Coot, and clicking a residue or atom will zoom in on that site in the graphics window. Coot will also display clashes detected by PROBE.
  • POLYGON: graphical comparison of user-selected model statistics with similar structures in the PDB.
  • PDB Statistics Overview: This very simple application uses the same database as POLYGON, but presents it in a different format. Useful when you simply want to view a histogram of a statistic for a filtered subset of the PDB.
  • Structure comparison: a tool for evaluating assorted model features and validation criteria for multiple related structures, and highlighting regions of difference. Automatically superposes model and maps into a common frame of reference for viewing in Coot or PyMOL.

We recommend including reflections in validation, since the model geometry ideally needs to be analyzed along with experimental data. However, the validation GUI will also run with just a PDB file. The phenix.refine GUI will also display a complete validation report at the end of refinement.

Utilities

  • Calculate map correlations: interface to phenix.get_cc_mtz_mtz.
  • Calculate model-map correlations: interface to phenix.get_cc_mtz_pdb.
  • Generate Table 1: utility for extracting statistics from PDB, reflection, and log files required for publication.
  • Diffraction image viewer: a simple utility for inspecting images in a variety of common detector formats (ADSC, MAR, R-AXIS, Pilatus, etc.).

Other tools

A number of simpler tasks involving file modification and visualization can be accessed by clicking the button labeled "Other tools" on the toolbar.

images/gui_misc_tools.png

Current options include:

  • Fetch PDB data: Download structure data in PDB, FASTA, or CIF/MTZ format. (Command line equivalent: phenix.fetch_pdb.)
  • Convert PDB file(s) to FASTA: Extract sequences of polymer chains in a model, with flexible options for handling gaps, insertions, etc. (Command line equivalent: iotbx.pdb_as_fasta.)
  • Multiple sequence alignment: uses the public-domain program MUSCLE (Edgar 2005) to perform alignment of sequences in various formats (including automatic extraction of sequences from PDB files). (Command line equivalent: none, but MUSCLE can be run in PHENIX with the command phenix.muscle.)
  • Plot average B-factors in PDB file: Plots residue B-factors (for all atoms, backbone, or sidechain) as a horizontal line graph. (Command line equivalent: mmtbx.bfactor_plot.)
  • Reindex reflections: Performs change-of-basis operations on reflection files. (Command-line equivalent: phenix.reindex.)
  • Modify PDB file: Shortcuts for performing some of the most common tasks in phenix.pdbtools (also available as a full GUI).
  • PDB remediation: Convert between versions 2 and 3 of the PDB format. (Command-line equivalent: iotbx.pdb_remediator.)
  • Import data and flags: Prepares data for use in refinement; this tool is especially useful when bringing in new high-resolution reflections for an already-refined structure.
  • Plot PDB statistics: Generates scatter plots or histograms of statistics output by phenix.model_vs_data for a large fraction of the PDB (nearly all X-ray structures with experimental data available). The same database is used to generate the plots in POLYGON.
  • Inspect R-free flags: Plots distribution of R-free flags in a reflections file; this is primarily used to determine whether the flags have been created in thin resolution shells. (Command line equivalent: wxtbx.inspect_r_free_flags.)

Projects

Like the CCP4 GUI (ccp4i), Phenix manages data and job history by grouping into projects. You will be prompted to create a project the first time you start the GUI. On subsequent launches Phenix will attempt to guess the project based on the current directory. There are several constraints on project management:

  • Project IDs should be no more than 24 characters and contain only alphanumeric characters (including underscore).
  • Project directories may not be nested, i.e. you may not create a project for a directory that is part of an existing project, or that has a subdirectory that is owned by another project.

When a project is created Phenix will create a folder ".phenix" in the project directory; this is used to store job history, temporary files, and other internal data. Users should not need to modify this folder unless deleting the project. All functions related to project management are available from the main GUI only, either in the toolbar or the File menu.

The current version of the GUI has limited functionality for configuring projects. Several options are available for setting the behavior of phenix.refine, such as default parameter files. The GUI will give you the opportunity to create these files automatically, but this feature is optional. The other primary function of the projects is storing job history; although this feature is still in development, it will save basic statistics and configuration files, and can restore the parameters and results of any successful previous run.

images/job_history.png

You may switch between projects while a GUI program is running; however, each program window stays associated with the project it was opened with.

Running PHENIX tutorials

A number of examples using published structures are included in the PHENIX installation, and can be automatically loaded into the GUI as new projects. To load tutorial data, click on "New project" in the main GUI; this window also appears the first time the PHENIX GUI is started (before any projects have been created). The button "Set up tutorial data" will open a new dialog with a list of examples, grouped by the methods they are intended to demonstrate.

images/tutorial_setup.png

For live demonstration, we normally use the P. aerophilum translation initiation-factor 5a or p9-sad example for experimental phasing, TEM-1 Beta-lactamase/beta-lactamase inhibitor complex (beta-blip) for molecular replacement, S. aureofaciens ribonuclease Sa (rnase-s) for refinement, and N-ethylmaleimide sensitive factor + ATP (nsf-d2-ligand) for ligand fitting. All of these datasets run relatively quickly (5-30 minutes) in the intended programs. (These examples are also used extensively in this documentation.)

Preferences

Some of the behavior of the GUI can be customized via the Preferences dialog, which is available from the File menu (Linux) or Phenix menu (Mac) and on the toolbar of most programs.

images/gui_prefs.png

Because some of the larger programs may have up to 500 distinct parameters, many of which rarely need changing, advanced settings are hidden by default. You may control this in individual dialogs using the "User level" menu at the bottom of the window, or change the global level in the "PHENIX interface" pane (shown above), in the "User level" option. For example, in phenix.refine, a typical configuration dialog will look like this when the user level is "Basic":

images/gui_level0.png

When set to "Advanced", many more options will appear:

images/gui_level2.png

Other Preferences settings determine what external programs are used for various file types, interactions with molecular graphics programs (Coot, PyMOL, and the simple built-in graphics), and options for some of the specific modules in Phenix.

Coot integration

Coot is an open-source (GPL) model-building program written by Paul Emsley. Although we do not distribute it with Phenix, it is available as source and binaries for Linux (from the developer's page), and third-party Mac binaries are available (provided by Bill Scott). Documentation on how to use it with Phenix is here. Phenix will try to locate Coot on your system automatically, but if it is not found, you may specify the command to use under Preferences->Graphics->Full path to Coot.

PyMOL integration

PyMOL is an open-source molecular viewer written by Warren DeLano. We distribute an older precompiled version (0.99) with Phenix; however, we recommend installing the latest version and add the command under Preferences->Graphics->Full path to PyMOL. More information is in the separate PyMOL instructions.

Reporting bugs

Bugs in the code may result in a pop-up window containing a detailed error message. Clicking "OK" will send an email to the Phenix developers with this information. We strongly encourage users to submit these reports, as they are one of the primary mechanisms by which we identify coding errors. The report will include the error message, information about the host system, and user name/email; no other personally identifying information or data is sent.

images/gui_error.png

Acknowledgments

The PHENIX GUI is written in Python 2.7, using the wxPython toolkit for most features, plus matplotlib for plotting. Some of the icons used in the GUI came from the Crystal Icons project which is licensed under the LGPL. Many additional icons (anything with a molecular representation) were generated using PyMOL.

Several open-source programs have been included with PHENIX and are used in various ways in the GUI; we are grateful to their authors for permission to redistribute the code:

  • **MUSCLE** written by Bob Edgar. Edgar, R.C. (2004) MUSCLE: multiple sequence alignment with high accuracy and high throughput.Nucleic Acids Res. 32(5):1792-1797.
  • **ksDSSP** written by UCSF Computer Graphics Laboratory. Original method (independently reimplemented): W. Kabsch and C. Sander, "Dictionary of Protein Secondary Structure: Pattern Recognition of Hydrogen-Bonded and Geometrical Features" Biopolymers 22:2577 (1983).