Protein 3D Structure and Structural Bioinformatics
Training by Eric Martz
Last updated March, 2008.

Seminars:
Two seminars are available. They have considerable overlap but different emphases.

Seminar Title:
Structural Bioinformatics and Protein Evolution Made Easy
Eric Martz, Professor Emeritus, Dept. Microbiology, University of Massachusetts, Amherst

Seminar Description for posters/emails:

Protein 3D structures are being solved at an accelerating rate, with expanding impact on understanding basic biological mechanisms as well as design of drugs. Researchers who are not specialists in crystallography or protein modeling need to be able to find, look at, and understand published 3D protein structures easily. I will demonstrate FirstGlance in Jmol (adopted by Nature among others), a popular free solution to this needs. I will address questions such as "What percentage of proteins have known structure?"; "What is the success rate of protein crystallography?"; and "How useful are homology models and purely theoretical models of proteins?" I'll touch on the status of the international Structural Genomics initiative. I'll demonstrate ConSurf, a completely automated server for identification and interactive visualization of conserved or highly mutated patches in protein molecules of known structure. Such patches are functionally important and can help to select residues for mutagenesis studies. Polyview-3D, the easiest way to prepare publication-quality molecular views and rotating Powerpoint slides, and Proteopedia, the easiest way to put customized molecular scenes online, will be shown.


Molecular BioSystems moves into a new dimension (anonymous, June 2006) and Added value online features for Molecular BioSystems by K. Sear describe the adoption of FirstGlance in Jmol by Molecular Biosystems (Royal Society of Chemistry) 2:345 (August 2006).

We're living in a 3D world (anonymous) describes the debut of FirstGlance in Jmol in Nature Structural and Molecular Biology 13:93 (February 2006).

ConSurf 2005: the projection of evolutionary conservation scores of residues on protein structures, Landau M., Mayrose I., Rosenberg Y., Glaser F., Martz E., Pupko T. and Ben-Tal N. Nucl. Acids Res. 33:W299-W302. 2005.

Protein Explorer: Easy Yet Powerful Macromolecular Visualization, Eric Martz, Trends in Biochemical Sciences, 27:107-109 (2002).

Other publications and bioinformatics resources that have adopted PE are listed at proteinexplorer.org/pe_lit.htm



Title for Workshop/Short Course:
Protein 3D Structure Visualization & Structural Bioinformatics Made Easy

Workshop Duration: A superficial whirwind tour can be done in half a day, but a full day workshop is minimal if novices are to leave truly enabled to find, look at, and understand 3D protein molecular structures. An optimal plan is to offer a second day for those who wish to practice skills introduced the first day, or learn about more advanced methods such as making animations, mutating residues in molecular models, or trying homology modeling.

Level of preparation required: Familiarity with the basic principles of the central dogma of molecular biology, and principles of protein structure will be assumed (at the level of a college biochemistry course). No previous experience with molecular visualization or modeling will be assumed. Appropriate participants are faculty, postdocs, staff, and graduate students in biochemical sciences.

Workshop Description for posters/emails:

Day 1: This hands-on computer workshop will enable novices to find published structures for molecules of interest, visualize their three-dimensional structures, and understand the resulting computer images. No previous experience with molecular visualization or modeling is required. The central tool will be FirstGlance in Jmol (adopted by Nature in 2006). In multiple-day workshops, Protein Explorer (2003 award for revolutionizing the teaching of biology at a molecular level) will also be employed. These tools easily highlight features such as backbone traces, disulfide bonds, secondary structure, ligands, and distribution of hydrophobic residues. Context-sensitive help is displayed automatically with each operation. Noncovalent bonds between ligands and receptors, between protein and DNA, etc. can be visualized in a few clicks. Key residues can be located by sequence number. Other powerful yet user-friendly servers will be introduced: ConSurf automatically colors amino acids by evolutionary conservation; specific oligomers will be constructed; customized, publication quality molecular images and rotating Powerpoint slides will be created easily with Polyview-3D; on-line molecular views in Jmol will be created with easy scene authoring-tools in the Proteopedia wiki; membrane planes will be visualized for integral membrane proteins; and model quality will be evaluated. Lesson plans, on-line tutorials, and resources for educators will be discussed. All software employed is free and works on Windows and Macintoshes.

These topics will be discussed, using slides available on-line for later use by all participants: Day 2: Those who attend the second day of the workshop may choose topics of interest such as creating animations (morphs of conformational changes), more depth on model quality analysis (with rules of thumb for interpreting resolution, R, and Rfree), mutating molecular models, and an introduction to homology/comparative modeling.

Protein structure experts familiar with other visualization software may wish to try out these features, many of which can be done in "one click": Visualizing noncovalent bonds between any two moieties, with checkboxes to hide/show water, water bridges, non-water hydrogen bonds, hydrophobic interactions, salt bridges, cation pi orbital interactions, and metal/misc. Displaying likely salt bridges and finding energetically significant cation-pi interactions throughout a protein. Seeing key information in the PDB file header in an organized, human-readable, clickable format (in Protein Explorer's Features of the Molecule control panel); Visualizing specific oligomer contacts (from EBI's Probable Quaternary Structure server); Animating functional conformational changes or movements, such as the binding of calcium to an EF-hand;

Computer lab setup
Windows / Apple Macintoshes / linux

Windows
  1. Any 32-bit Windows platform is suitable, including 98SE, NT4, 2000, ME, XP, etc.

  2. Java: Many resources in this workshop require an up-to-date version of Java. Windows XP systems, in particular, may lack the java virtual machine (JVM) completely, but installation is quick and easy from java.com. There you can also click the "Verify Installation" link to see if the java installed on your computers is up to date. If it is not, please update your java!

    For a one-day workshop, only items 1-2 above are required.
    For the second day of a two-day workshop, items 3-9 below are also required.

  3. Browsers: One of these browsers is required: Free Firefox version 1.5 or 2.0 is optimal for Protein Explorer. Download it from mozilla.com. Netscape 7.2 (not Netscape 8) can also suffice. Netscape must be version 7.2 because versions earlier than 7.1, and version 8 are problematic. Firefox and Netscape 7.2 (members of the Gecko family) are the only browsers that support Protein Explorer's MolSlide maker and Features of the Molecule control panel.
    • Internet Explorer: Although most of Protein Explorer works in Internet Explorer 6, neither the Features of the Molecule control panel, nor the MolSlide maker work in this browser. Protein Explorer does not work at all in Internet Explorer 7.
    • FirstGlance in Jmol works in Internet Explorer 6 or 7, and in Firefox and Netscape 7 or 8. The requirement for a Gecko browser is for Protein Explorer.


  4. The MDL Chime plugin must be installed in all browsers to be used for Protein Explorer. Test Chime Here. Right click on Chime, then click About Chime to see Chime's version. If your version of Chime is older than 2.6 SP6, please update it! The Chime installer puts Chime in Internet Explorer and Netscape 4, but manual installation is required in Gecko browsers. This is very easy -- you just copy a single file.
    Although MDL Chime is free, the MDL download site is burdensome. We recommend that you place a copy of the Chime installer on a local intranet server for convenient downloading within your institution.

  5. Try the Quick-Start link in Protein Explorer to verify that the browser is compatible. If you see a rotating 3D molecule, Chime is working correctly for use of Protein Explorer on-line. If you have problems, see Troubleshooting.

  6. Right click on atp.pdb then "Save Link As" to download a small PDB file. Drag the file atp.pdb and drop it onto a shortcut to the Firefox browser. The window should turn black, and you should see a wireframe rendition of ATP (shown at right). If, instead, if you get an "Open or Save File" dialog, then Firefox is not configured correctly to display downloaded PDB files in Chime. This will interfere with advanced uses of Protein Explorer, including the creation of MolSlides. To correct this, see step F.5. in Troubleshooting.

  7. Provided the lab computers have good Internet connections, it is not neccessary to have local copies of Protein Explorer on the lab computers, nor is it necessary to serve anything from a local server. Typically, workshops at diverse locations throughout the USA, in Spain and Japan have been able to use Protein Explorer from proteinexplorer.org without any problems.
    If for some reason local copies are desired, they are readily downloadable.

  8. Participants will each some personal disk space for saving files, such as PDB files and MolSlides. A few megabytes per participant should suffice.

  9. Optional: DeepView = SwissPDB-Viewer: the current version is will be useful in case some participants wish to do modeling, such as mutation of models, structural alignments, or homology/comparative modeling. However, it can usually be downloaded and run by participants if needed, since it does not require a formal Windows installation. Download DeepView.
Apple Macintoshes
  1. Apple Macs running OS X are feasible for a one-day workshop because their java can run Jmol-based resources. However, they are usually are not equipped to run Protein Explorer (see below), which would seriously limit what can be done in the second day of a workshop.
    • OS 9 can run Protein Explorer (see below) but cannot run the more important Jmol-based resources due to limitations of its java. Therefore, OS 9 is no longer suitable for these workshops.

  2. Java: OS X automatically includes java. Apple updates should be run to be sure the java is current. Current java is crucial for the many Jmol-based resources central to these workshops.

  3. The Safari browser is optimal for the Jmol-based resources central to these workshops.

  4. Participants will each need some personal disk space for saving files, such as PDB files and MolSlides. A few megabytes per participant should suffice.

  5. Optional: DeepView = SwissPDB-Viewer: the current version could be useful in case some participants wish to do modeling, such as mutation of models, structural alignments, or homology/comparative modeling. Download DeepView.

  6. Protein Explorer is problematic on OS X, but is unnecessary in a one-day workshop. If Protein Explorer is desired, the solution depends on whether your Macs are PPC or Intel.

      Intel Macs

    • On Intel Macs, Protein Explorer requires a Windows subsystem. (The Classic environment is not available for Intel Macs.)
      • The best implementation is Desktop for Mac (about $80) from Parallels.Com. Purchase this as a download and use the on-line installation instructions. Instructions and software boxed in stores tend to be quite obsolete!
      • Then you need to install MS Windows (at additional cost for a license). This gives you a window on the Mac Desktop that contains MS Windows. You have both OSX and Windows running concurrently (not an alternate boot) and you can use either as needed.
      • Finally, you need the Firefox browser with the MDL Chime installed (see instructions above).
      • This solution supports all capabilities of Protein Explorer, including the creation of MolSlides in Jmol that can then be viewed in Safari on Mac OS X.

      PPC Macs

    1. PPC Macs do not support the creation of MolSlides with Protein Explorer, but they support all other capabilities of Protein Explorer.

    2. Under OS X on PPC Macs, Protein Explorer requires the Classic environment (OS9). Classic is not necessarily included in newer releases of OSX, such as Tiger, so may need to be installed separately.

    3. The Netscape 4.7x/4.8 browser is required. Netscape 4.8 downloading instructions.
        (Netscape can coexist just fine with other browsers. Neither the Chime plugin nor therefore Protein Explorer can run in Safari or Firefox/Mozilla/Netscape 7-8 or Internet Explorer on Macs.)

    4. The MDL Chime plugin must be installed in Netscape 4 (Classic). Test Chime Here. Click on Chime and hold down the mouse until a menu appears, then click About Chime to see Chime's version. If your version of Chime is older than 2.6 SP3, please update it! (Although SP6 is available for Windows, SP3 is the most recent version of Chime available for Mac OS9.)
      Although MDL Chime is free, the MDL download site is burdensome. We recommend that you place a copy of the Chime installer on a local intranet server for convenient downloading within your institution.

    5. Using Netscape 4, try the Quick-Start link in Protein Explorer to verify that the browser is compatible. If you see a rotating 3D molecule, everything is working. If you have problems, see Troubleshooting.

    6. Provided the lab computers have good Internet connections, it is not neccessary to have local copies of Protein Explorer on the lab computers, nor is it necessary to serve anything from a local server. Typically, workshops at diverse locations throughout the USA, in Spain and Japan have been able to use Protein Explorer from proteinexplorer.org without any problems.
      If for some reason local copies are desired, they are readily downloadable.
Linux

  1. Linux systems are suitable for a one-day workshop because they support the Jmol-based resources that are central to these courses. For a two-day or longer workshop, a Windows subsystem is necessary in order to run Protein Explorer (see details below).

  2. The Mozilla Firefox browser is the only one that we can confirm works well for Jmol on linux.

  3. A current version of Sun java must be installed (java.com). Test your installation by going to firstglance.jmol.org, clicking on the Gallery, and loading a molecule. If you see the molecule rotating, you should be in good shape.

  4. Protein Explorer would be highly desirable for a workshop greater than one day in duration.

    1. Because Chime does not work on linux, neither does Protein Explorer, but it can run well in a Windows subsystem under linux. Among the solutions listed at this link, the only one I can confirm working well is VMWare. I have taught multiple day workshops quite successfully on linux systems with Windows available under VMWare.

  5. I have no experience with DeepView on linux, but it is "not supported" by the DeepView developers. The Windows version could be used in a Windows subsystem.