ABSTRACT:
--------
I am looking for experimentally-phased electron density maps of
proteins, to test an algorithm I have developed for calculating and
visualizing ridgelines of density (see below). This new algorithm
should be significantly more accurate and more robust than either the
Greer skeletonization or the critical-point graphs used in Grinch, but
of course I need to demonstrate that. For these tests I need maps
that were, or would be, used for initial chain tracing of a new
structure.
MORE DETAILS:
------------
I am working on an improved system for ridgeline visualization of
electron density maps, for use in initial chain tracing. It
involves a new approach to computing the shape of the density,
which is mathematically well-behaved and quite insensitive to grid-
spacing artifacts. I would like to test the system with a variety
of maps that come from the first stages of the interpretation
process, at different resolutions and quality levels.
This work is done within the GRIP Molecular Graphics project in
the Department of Computer Science at the University of North
Carolina at Chapel Hill under Fred Brooks and in collaboration
with Jane and Dave Richardson at Duke University. This system
will run on any reasonable platform with X-Windows, and its output
will be made compatible with major packages that do more detailed
fitting and refinement.
If you have a map or maps that you'd be willing to provide, please
contact me or just send it. I would need F's and phases (from
MIR, MAD, etc., with or without solvent flattening, but NOT ones
from molecular replacement or after any refinement), along with a
brief header or message with space group, cell dimensions, and
phasing method; reference to a PDB file for the coordinates; a
note about how the chain was in fact traced and how long it took
and/or how hard it was; and Email address for a contact person.
Thanks very much for any help you can provide, and I will let you
know the results.
CONTACT:
-------
Juraj Horacek
Department of Computer Science
UNC Chapel Hill, NC
Email: horacek at cs.unc.edu
