Structure visualization in biochemistry education.

Modern biochemistry and the other molecular biosciences have been fundamentally changed by the sequencing of the genomes of numerous organisms including humans. This genomic sequence information has led to structural genomics, which attempts to understand the function (or malfunction) of a protein from its predicted structure. Scientific journals and popular literature abound with illustrations of macromolecular structures illustrating these rapid advances. Today's students in widely diverse fields ranging from medicine to research (e.g. drug design) to teaching must therefore understand the relationship between structure and function, a relationship that has been called molecular 3D literacy (Richardson and Richardson, 2002).

Historically, chemists have used models of various types to render models of "invisible" molecules visible so that structural features can be analyzed and stereochemistry can be communicated. For example, in general chemistry and sophomore organic chemistry an understanding of the three dimensional nature of molecular structure can be obtained with hand models such as wire frame or hard sphere because the molecules are small enough to easily see, build, and manipulate. In the classroom and on examinations, the instructor typically asks students to draw 3D structures on paper using wedges, dotted lines, etc. to illustrate dimensionality. Some students can do this, but some have difficulty.

As structures get larger (> 500 Daltons), hand models and hand drawings become cumbersome, and for large macromolecules hand models are primarily of historic value or for illustrating small pieces of structure. There are a small number of free graphics programs in common use in biochemistry education to fill this gap. With ubiquitous laptop computers and data projectors, these programs can be used in lecture or laboratory, as well as in homework and online tutorials. I would like to briefly review these programs, (1) discuss their features, their strengths and limitations, and make some suggestions about how to use these programs most effectively in teaching biochemistry.

CHEMSCAPE CHIME

Today's biochemistry textbooks have illustrations of macromolecules often comprised of ribbon drawings to illustrate patterns of secondary and tertiary structure, or spacefilling models to illustrate shape and fit. Those textbooks also come with CDs and websites, which have interactive computer illustrations of these same molecules. These usually use Chime, a free Java-based web browser plug-in from MDL Systems (www.mdl.com), and the illustration is often imbedded in a webpage describing the illustration and providing buttons allowing specific changes to the graphic rendering. Likewise, when browsing the Protein Data Bank (Berman et al., 2000), Chime is the primary viewer for rendering the PDB files "visible" although there are other viewers available. Chime allows a certain amount of manipulation such as rotation, zoomingand switching from ball and stick to spacefilling rendering. Many general and organic chemistry teachers also use Chime because it is very good at rendering small molecules and has almost n o learning curve for simple rendering. As a web-based program, Chime is not plafform dependent. However, it is dependent on both browser type and version, with Netscape 4.7 being the most reliable of these. The next Chime release (version 2.6 SP4) should be fully compatible with Internet Explorer 6.0.

As mentioned above, there are a number of Chime-based macromolecular structure tutorials and many small molecule libraries available. Making your own Chime-based instructional tools is not trivial because one needs to be fairly proficient in HTML coding. Other limitations of Chime are that it is not open source and the graphics are only adequate for large structures, certainly not publication quality.

RASMOL

Chime is based on the stand-alone program Rasmol (RASter MOLecules), written by Roger Sayle in his spare time as he worked on his doctoral thesis in computer science at the University of Edinburgh. He made it public domain in June 1993 and it has gained a large following (Sayle and Milner-White, 1995; Bernstein, 2000). Rasmol has menu commands, but also has a command line interface, which allows much more manipulation of the image than Chime. It also has the nice feature of allowing images to be saved in a variety of formats suitable for presentations. Like Chime, Rasmol has almost no learning curve for simple rendering. For instructor or student to select features in Rasmol the user must learn the command line codes. These commands can be saved as scripts. Most students are not enthusiastic about using a command line and the scripts have the file paths embedded, making them difficult to transport to a different machine. In addition, Rasmol does not have depth cueing or the ability to load multiple molecules. An experimental Windows interface written by Philippe Valadon called RASTOP is designed to overcome these latter deficiencies.

PROTEIN EXPLORER

Protein Explorer (PE) is a web-based interface to Chime that, as the name says, encourages exploration of the structure (Martz, 2000). It was developed and championed by Eric Martz at University of Mass Amherst with NSF support. PE has lots of explanatory helps and there are good teaching ideas on the PE website (www.proteinexplorer.org). The interface also provides additional functionality such as a multiple sequence alignment coupled to structure overlay. PE is subject to the same browser limitations as Chime and tends to be limited by connection speed. The PE interface is powerful and provides a great deal of useful information, but the button interface can be very confusing. Finally, PE is in a constant state of revision, which is good for addition of new features but bad for stability.

DEEP VIEW

Swiss PDB Viewer, now called Deep View, is a stand-alone program that was developed by Nicholas Guex and coworkers at SmithKlineBeecham in Geneva, Switzerland (Guex et al., 1999) and championed as an educational tool by Gale Rhodes at the University of Southern Maine. This is an excellent graphics program for those destined for graduate school in biochemistry or related fields. Deep View serves as the graphics interface to the EXPaSY Swiss Model server where one can submit sequences for automated homology modeling. Deep View will load multiple structures, superimpose structures, do energy minimizations, and show a variety of surfaces and other renderings. It currently does not have depth cueing and the graphics are adequate, but they can be enhanced by exporting the image to PovRay. Deep View is more capable than the other programs reviewed, but also has a steeper learning curve than the other programs and so is not as widely used in education. Future versions of Deep View are planned with depth cueing and b etter scripting capability.

KINEMAGES

Kinemages and their associated programs, originated and maintained by David and Jane Richardson at Duke, constitute the original molecular graphics freeware. Written in 1992 for distribution with the journal Protein Science (Richardson and Richardson, 1992, 1994, 2002; Bateman et al., 2002), kinemages ("kinetic images") are plain text scripts which produce images when viewed with Mage, the kinemage viewing program. Kinemage scripts may be generated or "authored" with any text editor if the author follows the proper formatting but, for generating molecular kinemages from coordinate files such as Protein Data Bank files, there is a menu-driven authoring program called Prekin. The Prekin user interface is not particularly intuitive but is currently being rewritten to make it more user friendly. Kinemages have an extensive color palette and good depth cueing, so they lend themselves to artistic expression. Since they were designed to illustrate the viewpoint of the author, they are an excellent 3D communication tool. Kinemages are not limited to molecules and have been used for illustrations from geologic faults to architectural drawings to population data.

In addition to the programs briefly reviewed here there are other freeware graphics programs, which the reader might find useful. These include Echem, a simple program useful for high school and freshman college chemistry (Wu et al., 2001), MolMol (Koradi et al., 1996), and Cn3D (Wang et al., 2001), the graphics program available at the NCBI molecular modeling database.

EFFECTIVE USE OF MOLECULAR GRAPHICS

Educators must remember that factors other than biochemistry content are at work when using computer-generated molecular models. These other factors include the basic curiosity of the student when exploring a structure on their own, the difficulty experienced by the student in learning the software, and the spatial abilities and learning style of the students. For example, a visual learner may find molecular graphics more interesting than a student who has traditionally learned by reading the text.

My own experience in using molecular visualizations in lecture can be summarized in the following bullets:

* Use motion to give 3D effect (rock the image gently).

* Teach the basics of the software while you are using it.

* Make the images you use in class available outside of class.

* Give follow-up homework (graded!).

* Don't overdo it. Never spend more than 5-10 minutes on visualizations in lecture.

* Point out relevance of molecular structure to the function of the molecule.

* Use computer models jointly with solid models for small molecules.

* Keep the structure in overall context, i.e. together with sequence, physical properties, experimental data, biological role, etc.

Final Take home lesson: Make the students do some work. Active learning is essential. Passively watching colorful animated tutorials is little better than Saturday morning cartoons.

Summary Table

Graphics          Strengths                 Limitations
Program

Chime      Online toutorials,        Browser-dependent,
           platform-independent      relatively low quality
                                     images

Protein    Chime + extra helps       Same as Chime + always
Explorer                             in beta state

RasMol     Structure browsing &      Single structure, command
           exploration               line interface

Deep View  Excellent analysis tools  Steep learning curve

(1.) All programs in this article are linked to http://ocean.Otr. usm. Edu/~rbateman/hotlist.htm

LITERATURE CITED

Bateman, Jr., R.C., D. Booth, R. Sirochman, J.S. Richardson, and D.C. Richardson. 2002. Teaching and Assessing Three-Dimensional Molecular Literacy in Undergraduate Biochemistry. J. Chem. Ed. 79:22-23.

Berman, H.M., J. Westbrook, Z. Feng, G. Gilliland, T.N. Bhat, H. Weissig, I.N. Shindyalov, and P.E. Bourne. 2000. The Protein Data Bank. Nuc. Acids Res. 28:235-242.

Bernstein, H.J. 2000. Recent Changes to RasMol, Recombining the Variants. Trends Biochem. Sci. 25:453-435.

Guex N., A. Diemand, and M.C. Peitsch. 1999. Protein Modeling for All. Trends Biochem. Sci. 24:364-367.

Koradi R., M. Billeter, and K. Wuthrich. 1996. MOLMOL: A Program for Display and Analysis of Macromolecular Structures. J. Mol. Graph. 14:51-55.

Martz E. 2002. Protein Explorer: Easy Yet Powerful Macromolecular Visualization. Trends Biochem. Sci. 27:107-109.

Richardson, D.C., and J.S. Richardson. 1992. The Kinemage: A Tool for Scientific Communication. Protein Sci. 1992.1:3-9.

Richardson, D.C., and J.S. Richardson. 1994. Kinemages--simple macromolecular graphics for interactive teaching and publication. Trends Biochem. Sci. 19:135-138.

Richardson, D.C., and J.S. Richardson 2002. Teaching Molecular 3-D Literacy. Biochem. Mol. Biol. Ed. 30:21-26.

Sayle, R.A., and E.J. Milner-White. 1995. RASMOL: Biomolecular Graphics for All. Trends Biochem Sci. 1995. 20:374-376.

Wang Y, L.Y. Geer, C. Chappey, J.A. Kans, and S.H. Bryant. 2000. Cn3D: Sequence and Structure Views for Entrez. Trends Biochem. Sci. 25:300-302.

Wu H.K., J.S. Krajcik, and E. Soloway. 2001. Promoting Understanding of Chemical Representations: Students' Use of a Visualization Tool in the Classroom. J. Res. Sci. Teach. 38:821-842.