Here's a preliminary summary of the suggestions I got from the ccp4
community regarding the problem stated below (calculate theoretical SAXS
data from EM reconstruction):
The program em2dam, currently developed at EMBL-Hamburg (where the
magicians of SAXS live), converts a Spider map into a pdb file, which
can then be used as input for another EMBL-Hamburg program, crysol.
This approach does what I want. Nice. Em2dam hasn't been released yet,
but is available for testing upon request (ashku...@embl-hamburg.de).
http://www.embl-hamburg.de/ExternalInfo/Research/Sax/software.html
HYDRO can also calculate scattering form factors from dummy-bead models,
but I didn't try that out. From the group that wrote HYDRO comes
hydromic, a program that calculates, from an EM map, a distance
distribution and a P vs. h list whose usefulness didn't immediately
reveal itself to me. Those more familiar with SAXS might find that this
program is just what they want.
http://leonardo.fcu.um.es/macromol/programs/programs.htm
The program vol2pdb, which is part of the situs package, converts spider
format em maps to dummy-bead models. I couldn't suppress the feeling
that the pdbs obtained with em2dam were closer to the truth. Situs also
has a SAXS tutorial that might be useful.
http://situs.biomachina.org/tutorial_saxs.html
The EMAN program proc3d with the calcsf option can calculate scattering
curves. The output file contains the F**2/s distribution for the map
with max=0.0. Some scaling might be required. I didn't try this.
http://blake.bcm.tmc.edu/eman/
Chimera might have a SAXS function. I didn't check.
http://www.cgl.ucsf.edu/chimera/
Thanks to all who contributed ideas and suggested approaches, and
special thanks to Alex for sharing em2dam.
Andreas
On 23/07/2010 11:54, Andreas Förster wrote:
Dear all,
the other day I obtained SAXS data from which a low-resolution
structural model was calculated. The model is simpler/less complex than
one of the same protein that we obtained with cryo-EM.
Is there a way to estimate theoretical SAXS data from a cryo-EM
reconstruction to compare with the obtained raw data? Is there a program
that does for a reconstruction what CRYSOL does for pdbs? I understand
that there would be a huge amount of handwaving involved, but it might
help us reconcile our models.
Thanks.
Andreas
--
Andreas Förster, Research Associate
Paul Freemont & Xiaodong Zhang Labs
Department of Biochemistry, Imperial College London
http://www.msf.bio.ic.ac.uk