[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[ccp4bb]: summary: NCS analysis

***  For details on how to be removed from this list visit the  ***
***    CCP4 home page http://www.dl.ac.uk/CCP/CCP4/main.html    ***

Dear all,

just in case somebody (except for me :) is interested in extracting an
optimal 222 axes set out of a structure here is a summary of the replies
I got regarding my question:

Obviously, there is no easy answer to this problem. Nobody could point me
to a program that reads in a pdb-file containing four monomers with
non-crystallographic 222 symmetry and writes out three mutually
perpendicular 2-folds which best describe the approximate symmetry of the
tetramer in the pdb-file.

BUT, there are some 'approximations' which work as well, at least for the
time being and the goal in mind. 

1. Guogang Lu referred me to his website, which, besides some other useful
stuff, hosts a program called FIT:

FIT superimposes molecules (like lots of other programs do, too) AND
writes the rotation axis out as a vector (described by two points with
orthogonal coordinates) of arbitrary length, defined by the user. Then the
whole problem is reduced to lsq-fitting the three two-folds (= six points)
to another six points 'simulating' an ideal set of three two-folds with
the same length as was used in FIT. 

In this context Ed Berry pointed out that one could include the origin as 
a seventh point for fitting if the point closest to the three axes could
be determined. Which sounds reasonable to me but I haven't had the time
yet to check whether such a routine is available from somewhere or whether
one would have to write it from scratch.

2. Eleanor Dodson referred me to Kim Henrick who again recommended and
helped me with a no longer supported program in CCP4 - SYMFIT. 
Here it is possible to specify the complete symmetry you're
looking for in terms of symmetry operators and then put in sets of 4
corresponding C-alpha atoms (e.g.) from the four subunits. Which are then
fit/optimized by the program with respect to the specified symmetry

The calculated/optimized coordinates for the CA atoms can then be taken
into FIT again which gives three beautifully perpendicular axes.
Disadvantage is the limitation to 120 sites (the program was originally
written for heavy atom sites), so I can only put in ca. 1/10 of my CA
atoms. Maybe that problem can be solved by 're-dimensioning' the source
code, though.

3. Fred Vellieux also provided me with a program for superposition of
structures (SUPPOS). The operators obtained can be put into a subsequent
refinement/optimization based on a Bhat OMIT map (e.g.) I have to admit,
though, that I have not pursued this strategy yet since I wanted to avoid
working with maps - as long as possible.

Thanks to all who took interest in the problem!

Trixie Wagner, Ph.D.                                 phone: (804) 982-5764 
Dept. of Biology                                       fax: (804) 982-4896
Gilmer Hall
University of Virginia
Charlottesville, VA 22903