BIO/CSC295 2009F, Class 18: Structure Prediction (1)

Admin:
* Fun event Saturday a.m.: Grinnell Destination ImagiNation Fun Run/Walk.  
  * More info outside my office.
* If you are registered in Powesheik county, don't forget to vote.
* Don't forget Thursday's "Thursday Extra": 
  * The Ontological domain model of medical imaging informatics.  
  * (Yes, it will be more interesting than it sounds.)
* Famous Grinnell alums: Kumail Nanjiani '01 made the front page of the
  NYT Sunday A&E section.
* Looking ahead: The environment and human rights symposium next week,
  including former PETA member (?) David Cantor '77.
* Calendar software being upgraded soon, letting you more easily search
  and add to your Google calendar and ...

Overview:
* Protein structure: What is it and why care?
* The Chou-Fasman Algorithm
* Web exploration
  * PSIPRED stuff available in Examples directory (add to end of
    Course Web site)
  * Run Cn3D with /home/rebelsky/bin/cn3d

Protein Structure
* If you took good notes, Vida talked about it already
* And if you did the reading, you'd know a lot, too.
* Four levels: Primary, Secondary, Tertiary, Quartenary
  * Primary: The sequence
  * Secondary: Formation of ordered structures (the main topic)
    alpha helices and beta strands, which form beta sheets
  * Tertiary: Interactions between amino acids to produce overall
    3D fold.
    * "When you think about a generic protein, what does it look
      like in your head?"
      * A new quote for Chris French's plan: F says "A hairball"
      * Also: Tumbleweed, a Pacman (registred trademark of ...)
    * But there are also many different appearances
      * E.g., trans-membrane channels, beta barrel (like a tube)
        * Permit transport of specific molecules
  * Quartenary: Combination of chains
* What causes proteins to look the way they do?
  * Fold is the lowest energy state for a specific amino acid sequence
    * Caveat: Constructed from end terminus to carboxy terminus
      * And some things begin to fold being constructed/translated
      * So it may be locally optimized lowest energy state
    * Strutures sometimes fold and unfold while being made
    * Associated proteins (chaperone proteins, etc.) may help in the folding
  * Things that affect folding: Size, hydrophobicity, etc.
* Prediction of protein folding is a VERY DIFFICULT problem
  * Theoretically, you should be able to calculate it if you understand
    all the parameters correctly
  * We can talk about propensity to form alpha helix or beta sheet.
  * The problem interests lots and lots of people.
* If computationally predicting the 3D structure of a protein is difficult,
  how do we end up with such structures?
  * X-ray crystallography
    * Convince the proteins to form a crystal
    * Shine an x-ray through the crystal
    * Look at the diffraction pattern
  * Linus Pauling did structure of alpha helices
  * Rosalyn Franklin did some of the first really good structures
  * Dorothy Hodgkin did the structure of penicillin
  * People have also used NMR and electron microscopy
  * E.g., tubelin is hard to study through x-ray crystallography because
    it polymerizes rather than crystallizing
    * So people used electron microscopy
* Goal: Can we get information about the structures?  (And why do we care?)
  * Drug targeting
  * Looking at interactions between models
  * Confirmations of what binds to the protein and where
    * Helps you understand the function of the molecule
  * Understanding of mutations
  * Note: A student in this class studied these kind of issue with respect
    to neurotransmitters
  * Structure can also give you a sense of what the protein might do
* Note: A different model of drug design
  * Much of current drug design is "Hmmm ... this seems to have an effect;
    let's see how to modify it."
  * The model the book is discussing is more of a "Okay, here's what the
    protein looks like, let's build somethign to target it."

* Side note: RNA has 3D structure, too
  * tRNAs have very distinctive structures
  * Not quite as difficult as proteins, since pair in natural way
  * Form stems and loops and such
  * mRNA, tRNA, and cool regulatory RNA molecules
  * Bioinformatics can help us study these kinds of molecules
  
* Today we look at some of the tools available
  * For predicting secondary structure