X-Ray Crystallography
These notes are from the Spring 1995 Structural Biology class
given at Berkeley.
Note: These are incomplete and I have no intention of completing them.
X-Ray Chrystallography Table of Contents
part1
part2
part3
Back to the Table of Contents
The Phase Problem
Diffraction experiments collect only part of the data needed to determine the crystal structure of a protein. The following methods have been used to reconstruct the phase:
- Isomorphous replacement
- Introduce (by diffusion) new heavy atoms into the crystal, all of which bind in the same place on the protein
- Assumes that the new atom does not significantly perturb crystal structure
- A single Isomorphous replacement limits each phase to 2 choices-- a second isomorphous replacement can be done to narrow it down
- Molecular replacement
- Calculated from related structure
- Introduces bias towards the related structure
How Accurate is Crystallographic Data?
- Refinement:
- R value between 0.2 and 0.25 indicates that the chain tracing is most likely accurate but some side chains may be off
- R value below 0.2 indicates that there are no major problems with the structure, although some side chain errors may remain
- Without refinement, positional errors can approach resolution/3.
- A resolution of less than 2.5 Angstroms is largely correct
- A resolution between 2.5 and 3.5 Angstroms will be approximately 80% correct if the sequence is known
- A resolution of above 3.5 Angstroms indicates that errors may even exist in the chain trace
Back to the this-section Table of Contents
Last revised: 1995 May 12 by
sev@byzantium.mckusick.com