User:Eric Martz/Introduction to Structural Bioinformatics I

From Proteopedia

Revision as of 15:09, 29 September 2012

Overview and Goals

In these two classes, you will

1. Be introduced to structural biology and structural bioinformatics.
2. Choose a protein molecule to work with (different for each student).
3. Learn how to use easy tools to visualize answers to questions about the structure of your molecule. The main tools are:
   1. Proteopedia.Org, a wiki encyclopedia of protein structure.
   2. FirstGlance in Jmol (FirstGlance.Jmol.Org), making it easy to see key structural features of a macromolecule.
4. Take snapshots of your molecule, and paste them into Powerpoint slides, along with your answers.
5. Email your .pptx file to emartz@microbio.umass.edu for grading.

Computer Lab Preparation (BCRC)

1. Log in
2. Run Firefox
3. Go to proteopedia.org (do NOT type www).
4. If you see inactive plug-in, click on it and ENABLE the plug-in.
5. Restart the browser and again go to proteopedia.org.
6. When you see a rotating 3D molecular structure, you are prepared.
7. Take a look around Proteopedia. Click on the PDB codes below, or the Random links to see other molecules.

Protein Structure and Structural Bioinformatics

1. Amino acid sequence + protein chain conformation = protein function.

A. Conformation can be a stable fold or intrinsically unstructured. Both commonly exist in the same protein molecule.

B. Conformation is specified by sequence.

• Folded domains fold spontaneously (Anfinson, 1960's^[1]), or with the help of chaperonins.
• The denaturation (unfolding) of a folded domain destroys its function.

2. Structure Knowledge.

A. Although sequence specifies fold, scientists cannot yet predict the fold from the sequence. Therefore, fold must be determined by empirical (experimental) methods. The most common methods for determining the 3D structure of a protein molecule are:

• X-ray crystallography, 88%.
• Nuclear magnetic resonance (NMR) in aqueous solution, 11%.

• NMR is limited to small proteins (30 kD or smaller).

• High resolution cryo-electron microscopy, 0.5%.

B. These methods are difficult and expensive. Less than 10% of proteins have known structure.

C. All published, empirically determined 3D macromolecular structure models are available from the Protein Data Bank (PDB; pdb.org; About the PDB).

D. Each model has a unique, 4-character accession code called a PDB identification code, for example

• 1hho
• 1d66
• 2bbn
• 9ins

Choose a Molecule to Explore

• Choose a molecule that includes protein and ligand. It may also include nucleic acid, but must have protein and ligand.
• Be sure to note the 4-character PDB code of the molecule you choose. The PDB code makes it easy to retrieve the molecule and information about it. Here are some ways to find a protein with known structure:

1. Pick a molecule in the Atlas of Macromolecules (Atlas.MolviZ.Org).

Choose a "straightforward" molecule that has ligand.

Notes and References

1. ↑ For a brief overview of Anfinson's protein folding experiments in the 1960's, see the first paragraph at Intrinsically Disordered Protein.