Model Building/Graphics
Structural Molecular Biology Laboratory, ChemM230D

A model building session using the graphics program "O".
Suggested Reading Materials

1) Electron density map intepretation by T. A. Jones and M. Kjeldgaard, Methods in Enzymology, Vol 277 (1997) pages 173-207.  This link is a 17 page PDF file 

2) Model building tutorial by Bernhard Rupp.

3) Model building introduction in powerpoint format by Mike.


Assignment & Procedures
4th Assignment: 
Model Building Report
due at the end of lab.
Objective: To illustrate the fold of your molecule using a ribbons diagram generated by a commonly used graphics program. 

Method:  Orient your molecule so that all secondary structure elements are visible and can be clearly labeled without confusion.  Print out the figure and label the strands numerically.  Label the helices alphabetically.

Ribbons representation of proteinase K structure generated using the graphics program "Ribbons" written by Mike Carson.  Beta strands are collored green.  Alpha helices are colored cyan. 310 helics are colored blue.  The structure was built using ARP/wARP automated chain tracing program.  The program requires only the structure factors and phases output by DM and the primary sequence of proteinase K.  The structure pops out after about 1/2 hour of calculations.

Part One: Modeling an alpha Helix.
Objective: To orient an alpha helix properly in electron density.

Background: Recent advances in crystallographic software have made it possible to build a protein chain entirely automatically by the program ARP/wARP if the native data set exceeds 2.3 Angstrom resolution or by the program "MAID" if the data exceeds 2.5 Angstrom resolution.  In these cases, the job of the crystallographer is reduced to that of tying up loose ends left by the refinement program.  If however, the native data is below medium resolution, the chain trace must be performed manually by the crystallographer.  One generally begins by modeling  secondary structure elements (i.e. alpha helices and beta strands).  Coordinates of idealized helices and strands may be downloaded and read in to the graphics program O.  The structural elements may then be grabbed and dragged into electron density.  

Procedures Type the word "ono" to start the "O" graphics program. Press the "enter" key until the graphics window appears. In the graphics window type @dmacro.  This instruction file will place a polyalanine helix in the middle of a clock face.  Use the grab group command to rotate the helix so that the c-terminus of the helix points to 3 o' clock.  Then translate the helix into a box on the right of the screen.  Once you have mastered the rotation and translation of protein fragments, you may proceed to fitting atomic models into electron density.  View of an alpha helix down the helical axisIn the graphic window type "@mbmacro".  This command will read in a poly alanine starting model and a few libraries that will aid in model building.  Tear off the following sub-menus form the "menus" menu: Objects, User, Fake Dials.  You should be able to see three objects: the poly alanine model (prok1); the solvent flattened electron density (dm); and the unit cell outlined in red.  You will see two parts of the density map which contain no model.  One area requires a helix, the other area requires a beta strand.  Begin with the alpha helix. Select from the menu REBUILD , GRAB, GRAB GROUP, the click on the helix.  Drag the helix model toward the corresponding density.  Look at the helix density from three angles.  First look at it lengthwise; adjust the helix to fit the electron density throughout the length of the density.  Change your view by 90 degrees and adjust the alignment again if necessary.  When you are happy with the placement, select YES from the CONTROLS menu.  This will save the new orientation of your helix.

"O" graphics session.  Cyan is bones, yellow is a model helix.
Introduction to "O": The task of model building into electron density maps is most frequently done with the aid of the graphics program "O" written by Alwyn Jones.  The program contains a number of features than enable the user to build a model in compliance with the known rules of amino acid geometry.  For instance, when building side chains into a medium resolution map, "O" can suggest choices amon energetically favorable rotamers to fit a given electron density map.  Also, when building an extended loop, the user may simply chose from a library of loop conformations collected from high-resolution structures.  Despite these powerful features, the program does contain a number of irritating buts and the syntax of commands is often non-intuitive.  The latestcomplete manual for O was written for version 8.  Currently we are using version 8.  There are also somerelease notes.
Controls for rotating and translating groups of atoms in O.

Correct helix orientation.  Carbonyl oxygens fit in density.

Incorrect helix orientation. Carbonly oxygens don't fit in density.
Part Two: Modeling a Beta Sheet
Objective: To fit a beta strand in the proper orientation.


View of beta strand showing the characteristic zig-zag pattern.  This view is most helpful for  initial positioning the beta strand model because it is very clear where the C-Betas should be pointing (i.e. into the side chain density). The strand runs N-term to C-term from left to right.

The same view as the panel on the left, except the model beta strand has been positioned with the strand running from C-term to N-term as it runs across the image from left to right (i.e. in the opposite orientation from the left panel).  As you can see, from this perspective both orientations (N to C or C-N) appear to fit equally well.

View is rotated 90 degrees from the image above about a horizontal axis.  The carbonyl oxygens fit well. This orientation (N-to- C) is the correctorientation.

Same view as on the left panel, except the model is oriented (C-to-N). The carbonyl oxygens do not fit the density.  This is the incorrectorientation.
Part Three:Placing the sequence.
Objective: To assign a sequence to your poly alanine model

 Fit about 10 amino acids in a row.  Compare with sequence (Tritirachium album) on right.  Once the proper registry is found, continue mutating according to the sequence on the right.  Use stereo glasses.









Representative electron density for amino acid side chains arranged in order of increasing size. 
From an experimental electron density map calculated at 1.5 Angstrom resolution.

Instructor's preparations

Back to CHEM M230D course syllabus 

[Overview] ·[Facilities] · [People] · [Services] ·[Lectures] · [BioLinks] · [Stats] ·[Search]