I hate this problem - the only real guarantee you have it right is when
the correlation coeffs are reasonably high..
The CCP4/MR programs/LSQKAB/ DM etc conventions are consistent I believe
{R11 R12 R13 ] [XoA] [XoB]
[R21 R22 R33] [Y0A] = [YoB]
[R31 R32 R33] [Z0A] [ZoB]
Omat always prints/reads the transpose of that rotation matrix.
Transpose is the same as inverse, which means that if one matches B to
A rather than A to B you get the Omat which adds to the confusion..
Eleanor
Huiying Li wrote:
*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
I was running DM with 3-fold NCS averaging on a set of MR solutions by
PHASER. I have generated the NCS rotation matrix and translation to
bring molA to molB or molA to molC using either LSQMAN or
LSKAB(through CCP4i GUI 1.3.20 and ccp4-5.0.2). The NCS averaging
worked (final correlations are > 0.7 for NCS related molecules) only
if I used Euler angles to define the rotation/translation. When I used
the rotation matrix, either ROTA MATRIX or OMAT, the best
correlations were ~0.18 or ~0.4, respectively. I had the same
problems no matter running DM through CCP4i or with the script. It
seems DM has problems in interpreting the rotation matrix.
A related issue is the rotation matrix convention. I recall the
matrices given by LSQMAN and LSKAB were used to be transposed to each
other. But this time I found they were identical for the same NCS
transformation. Has one of these two programs changed its convention
lately?
Thanks for any help.
-------------------------------------------------
Huiying Li, Ph. D
Department of Molecular Biology and Biochemistry
Natural Sciences I, Rm 2443
University of California at Irvine
Irvine, CA 92697, USA
Tel: 949-824-4322(or -1953); Fax: 949-824-3280
email: [EMAIL PROTECTED]
--------------------------------------------------
begin:vcard
fn:Eleanor Dodson
n:Dodson;Eleanor
email;internet:[EMAIL PROTECTED]
tel;work:+44 (0) 1904 328259
tel;fax:+44 (0) 1904 328266
tel;home:+44 (0) 1904 424449
version:2.1
end:vcard
