Largest protein structure determined to date by traditional ab initio direct methods
What is it about?
The nominal size limit for protein structure determination by direct methods (probability theory) is 1000 non-hydrogen atoms. The presences of a heavy atom enabled structure determination from the native diffraction data and probability theory as encoded in SIR2004. No anomalous difference data nor prior knowledge of the position of the heavy atom were used in structure determination. The structure of bacteriophage P22 was unknown and had defied structure determination for over 20 years. Sequential additions of glycerol to the reservoir after crystallization drop setup increased the volume of the crystal eight-fold and thereby led to the required 1.05 Angstrom data.
Why is it important?
We call this approach augmented direct methods. This is important because it shows how the purposeful addition of heavy atoms can extend the application of direct methods to larger proteins. This approach is an option when anomalous diffraction data differences and good molecular replacement models are absent. The results of a series of structure determination trials with truncated diffraction data demonstrated the weakness of the R-factor for detecting correct structures at resolution limits below 1.3 Angstroms. The figure that summarizes these trials is very useful in teaching, especially when emphasizing that many phase sets corresponding to random atom positions can lead to low R-factors and these R-factors decline as the resolution cutoff declines.