Biomedical Engineering Reference
In-Depth Information
In addition to Cn3D, RasMol, PyMol, SWISS-PDBViewer, and Chimera, there are dozens of alternative
free and commercial rendering systems available. Most of these programs tend to be predecessors of
more functional programs either under development or recently released. Even though many of these
programs overlap each other and the programs discussed here in functionality, they are typically very
good at a particular function and require less hardware power than the latest do-everything
packages. For example, MolMol supports the display of electrostatic potentials across a protein
molecule. MidasPlus, a predecessor to Chimera, has a sequence editing feature that allows the user
to define a point mutation—substitute one amino acid for another—and visualize the result on the
protein structure. MolSript, Ligplot, and Dimplot are specifically designed to create images for print
publication, in that they support fine control over the output formatting. Ligplot additionally
generates 2D schematic drawings of bonds and structure, and Dimplot, a variant of Ligplot, renders
interactions among multiple protein chains.
General-purpose rendering systems can be used to obtain images that fulfill special criteria, such as
extra high-resolution images for publication, enhanced color or transparency options to emphasize
specific regions on the protein, and other, custom applications. For example, a general-purpose 3D
rendering engine such as Strata3D Pro, Bryce, Max3D, or LightWave can render molecules within the
context of cell wall or other structure as part of an illustration for print or film production. Similarly, a
fly-through of protein structure can be rendered in one of these programs for teaching purposes.
Aside from cost, which can range from a few hundred dollars to several thousand dollars, the
downside of using one of these commercial, general-purpose rendering programs is that setup time
may be dozens of hours per molecule. Most of this time is typically spent translating data from PDB
or MMDB into a format supported by the rendering engine.
One of the challenges of working with multiple rendering engines is that the expected file format and
contents may vary from one system to the next. Most rendering applications, including RasMol, are
compatible with the PDB format, which contains a simple description of amino acid sequences.
Programs that use the PDB format are required to use rules—which may vary from one program to
the next—to construct the protein structure, based on sequence data alone. That is, these programs
not only render the image, but perform modeling of the underlying data as well.
The result is potentially wide variations between displays of rendering systems using the same PDB
data. In contrast, the MMDB contains data on the molecular bonds in the protein structure in ANS.1
(Abstract Syntax Notation number One) format. Because the data on molecular bonds is provided,
and not generated by the rendering engine, programs that render ANS.1 data tend to produce results
that are highly consistent with each other. This doesn't necessarily translate to greater accuracy in
the rendered image, however.
In addition to the PDB and ANS.1 formats, another common format for protein structure rendering
engines is mmCIF (Macromolecular Crystallographic Information Format). Note that the PDB supports
both PDB and mmCIF formats, as shown in
Figure 5-14
. Because reading in data in the mmCIF
relational format is so extensive, it generally requires too much in computational resources for
continual use. As such, it's much better used as an archival format. In contrast, data in the
hierarchical ANS.1 format loads quickly, which is one reason why the ANS.1 format supported by the
MMDB is preferable for viewing applications designed for browsing 3D protein structures.
Figure 5-14. Protein Structure Rendering Formats. Protein Data Bank data
for Glutamine Synthetase is available in PDB and mmCIF file formats.
