Pick two proteins of known structure and find something interesting to illustrate in their structure. I want two quite different proteins, with different stories to tell, not two closely related ones.

You can pick up pdb files directly from the PDB web site at http://www.rcsb.org/pdb/, or you can use the pdb-get script in /projects/compbio/bin/ to fetch the file to /projects/compbio/data/pdb/ and return the file name. It is probably worth your time to explore the PDB web site, particularly the search engine and the molecule of the month. The pdb-get script does automatic download from the PDB web site for you, but you need to be in the "protein" group, in order to have write privileges in the /projects/compbio/data/pdb/ directory.

Write text and make pictures that illustrate your point clearly. Do not take pre-existing images or ones created by a web site, but create your own picture.

Reading several of the "molecule of the month" descriptions on the PDB web site may give you some ideas what to look for in the structures. It could be a static feature, such as an active site or where the protein interacts with another. It could by a dynamic feature, such as a conformation change when binding a ligand (this can often be seen by superimposing structures solved with and without the ligand).

One popular research task these days is to map disease-related SNPs (single nucleotide polymorphisms) onto the structure of the protein that is the gene product, to try to predict how the SNP might disrupt the function.

Another possibility is to superimpose homologous proteins to highlight similarities and differences.

I'm not expecting you to discover anything new about these proteins, but to learn something interesting about them and to make illustrations with accompany text that conveys the story clearly to others. Two informative pictures for a protein are much better than five uninformative ones. And don't forget that you'll need explanations to accompany your images.

I'd also like a short review of the tool that you decided to use. Why did you pick it? How hard was it to use to get the pictures you wanted? What features did you find particularly nice? particularly frustrating?

This year, there are two variants on the assignment, and you can pick either one:

Variant 1: Static HTML file

To examine files from PDB, you will need one of the many visualization tools available. I still use Rasmol (available for free and running on many platforms), but there are many other fine tools available. Many people find that pymol, though having an awkward-to-learn interface, produces the best pictures. Both rasmol and pymol are installed in /projects/compbio/bin/i686, which you may need to add to your UNIX path. (See my .cshrc file for one way to add directories to your path.)

Other tools that may be worth considering are vmd (used in the molecular dynamics community) and chimera (developed by UCSF). Both of these claim to be highly extensible. It appears that vmd is installed, but not chimera. The installation instructions for chimera look pretty simple, and I've not looked at the vmd instructions yet.

Pick one or more of these tools, get access to a copy, and learn to use it. The programs rasmol, pymol, Cn3D, and molmol are installed on School of Engineering machines in /projects/compbio/bin/i686, as is Swiss PDB viewer (as deep-view), though they may not be the latest versions. You should learn how to rotate, zoom, and translate the molecules, how to get different drawings (at least cartoon, wire frame, and space-filling), how to select individual residues, and how to output images in various formats. Most of the molecular graphics programs have home pages with sample images that demonstrate their capabilities.

Hand in this assignment by making a static web page (no plugins needed and working with almost any browser), and sending me the URL. On the School of Engineering machines, you can make web pages by creating files in the .html/ directory under your home directory and making them publicly readable. JPEG and GIF are the preferred export formats for images on web pages, though PNG and SVG now seem to be almost as accepted.

The web-page format will limit you to somewhat simpler images than you might use in print, since the resolution is a bit limited. This may actually help you, since there is a tendency toward too much clutter from unnecessary details.

Variant 2: Proteopedia page

For this variant, you'll have to learn how to write JMOL scripts to make your pictures, but the pictures can be movies that zoom in and highlight particular features, which can be easier to make comprehensible than static images.

A major advantage of this approach is that you'll have a genuine audience for your work, which tends to improve quality.

If you do this variant, send me the URLs of the two proteopedia pages you edited. I should be able to tell from the history of the pages which of the work is yours (so you may need to tell me what name you registered with Proteopedia).

SoE home

Kevin Karplus's home page

Biomolecular Engineering Department

BME 220 home page

Karplus's lab page

UCSC Bioinformatics research

Questions about page content should be directed to