Statistical Phasing
The phase problem in X-ray crystallography has recieved a great deal
of attention from any number of gifted scientists over the last 50
years, and much progress has been made in providing partial solutions
for limited cases. For example, small structures are now solved
automatically and routinely by direct methods using the measured
intensities only, and large structures can be solved from
comparitively weak phase information by the application of density
modification. Direct methods have failed to solve any but the smallest
proteins, and density modification has a number of theoretical
problems which will probably prevent much further improvement of the
method (Cowtan & Main (1995), Acta Cryst D51).
I have become convinced that the only long term solution, both for the
improvement of current methods, and for the eventual solution of the
phase problem, lies in a proper statistical treatment of the phase
problem. This work has been pioneered by Gerard Bricogne in his
groud-breaking paper (Bricogne (1984) Acta Cryst A40, p410-445).
Unfortunately this revolutionary work has not recieved the attention
it deserves, partly due to the difficulty of the subject material. I
am therefore including here some of my own annotations, in the hope
that others may be encouraged to explore this exciting work. Expect
this document to grow...
Annotations to Bricogne (1984)
Back to my homepage.