*** For details on how to be removed from this list visit the ***
*** CCP4 home page http://www.ccp4.ac.uk ***
> It this point that I had in my mind. I would like to account simultaneously
> for_all_ symmetry-equivalent intra-molecular Patterson vectors at once. Since
> the intra-molecular Patterson vectors of all symmetry copies of the true
> structure fall into the same sphere around the origin, any true orientation of
> the search molecule will have its P1 Patterson self-vectors overlapping both
> with the self-vectors of the corresponding true molecule, which will add to
> the signal, _and_ with the self-vectors from its symmetry copies, which will
> add to the noise. Fitting them all simultaneously by applying the rotational
> symmetry of the space group to the P1 self-vectors of the search molecule
> would result in a Patterson overlap with all self-vectors of the true
> structure at once. This should enhance the signal-to-noise ratio of the
> orientation search.
> What do you and others think about this?
Apologies for the unashamed plug, but multidimensional molecular
replacement search methods (like Qs :-) do exactly what you described
[plus more, since (in the vector-space-based metaphor) they simultaneously
try to match all intra- and inter-molecular vectors, with the full
crystallographic symmetry explicitly accounted for]. On a related side
note, I believe that the major source of problems with the results from
the cross rotation function is not overlapping self-vectors, but
'overlapping' cross-vectors.
Nicholas
--
Dr Nicholas M. Glykos, Department of Molecular
Biology and Genetics, Democritus University of Thrace,
Dimitras 19, 68100 Alexandroupolis, Greece, Fax ++302551030613
Tel ++302551030620 (77620), http://www.mbg.duth.gr/~glykos/
