Voluntary Assignment 3A

Main Index...dated 23rd.Feb'95

Model Building

Molecular Models from Indigo Instruments.

However good computer graphics are, physical models are still extremely valuable learning aids. Remember the part that models played in the DNA story and in the solution of the early X-ray structures of proteins. We therefore strongly encourage you to think about building a physical model of your protein - (mainly just because it's fun!) It's also a most exciting thing to have around the lab or home.

We can't easily tell you how to do this, other than give some guidelines. A number of companies supply kits but the ones that most people use are Nicholson models - I can't give regional suppliers - I'm sure course members can. These consist of plastic units: peptide, carbons, functional groups (for the 20 aa) which click together. I'm *guessing* that it costs about $1 per residue so that it is affordable for most labs but not affordable for poor students (or others :( It takes about 1 month elapsed time to build a protein of about 250 residues, but at the end of it you will *really* have a feel for structure. I'm suggesting you start thinking *now* as it takes a little while to order the bits. You'll know enough about secondary and tertiary structure by the time the kit arrives.

If you're still reading, decide whether you want to build *your* protein or play safe and buy a myoglobin/lysozyme/insulin or other standard molecule. (You get specific instructions with those, including projections onto a baseboard, etc.) If you build your own, you'll have to count the aa's and order them, get a perspex (or other) baseboard, and some support rods. Also some H-bond links. Then decide which way up your protein is to go, and work out how it projects onto the baseboard. (Rotate the molecule on the graphics and then write out the transformed coordinates. Then just pick (say) X and Y - you'll have to work out which.) Select a number of key points where the structure should be physically supported and buy enough support rods for this (probably ca. 1 for every 15-25 residues.) Work out where (X,Y) the holes should be drilled in the baseboard and how long the rods should be (Z).

Building is normally done in torsional coordinates, so you'll need a list of the phi and psi angles of your backbone, and also chi1, etc for the side chains. There is usually a protractor with the kit, which allows you to 'dial' the correct torsions. You'll find your own way, but normally you would build a single residue, add the side chain, dial its correct conformation, add the next residue, dial phi and psi, and so proceed. I normally work with the secondary structure elements (e.g. make the helices first). By using the graphics you can visually check whether your chunk looks like the picture! You may have already drilled the support holes (e.g. in the c-alphas) before you started. Some people might try to assemble the whole chain; others might build the core and add the loops at the end. You might also find that your particular protein would benefit from an extra framework of some sort. As you go along you'll add the H-bonds and the disulphides. In any case, you'll probably find general instructions which the kit or someone will post them.

I really recommend it - post any feedback about kits, experience, etc.