I guess that what you are describing is a case of "local" symmetry, i.e.
it is not true symmetry in that it does not hold for the entire particle.
If this is the case, then you should run a refinement in tetrameric
symmetry (like you did), then cut out the 3 pseudo-symmetrical subunits
from the unfil.mrc maps, and then use some 3D program to align and average
over these 3 subunits. This will give you a further improvement in
resolution for your subunits. Similar procedures are used frequently for
T>1 icosahedrl viruses. One example that jumps to my mind is the rotavirus
structure from Niko Grigorieff.
BTW: Estimating a "gold-standard" FSC for the locally-averaged half-maps
may become a bit more difficult due to the masks involved in the local
> I'm working on a protein that has a tetrameric symmetry (symmetry: T).
> Each face of the tetramer features an additional C3 symmetry.
> Thus each particle has 12 asymmetric units.
> My initial runs with symmetry=T (4 asymmetric units per boxed particle)
> where very successful, but now I'm interested in applying the whole/right
> symmetry during the refinement.
> Is there a symmetry option (or hack) that fulfils my needs?
> Thanks and best,
MRC Laboratory of Molecular Biology
Francis Crick Avenue, Cambridge Biomedical Campus
Cambridge CB2 0QH, U.K.
tel: +44 (0)1223 267061