[MMTK] point group symmetry

Margot Ernst margot.ernst@univie.ac.at
Wed, 5 Jun 2002 12:40:28 +0200


Hi again,
I continue to be puzzled with the (lack of ) symmetry in my domain analys=
is.=20
Let me try and describe the problem:=20
I have a protein with subunits (aka chains) arranged symmetrically around=
 an=20
axis - i.e. a symmetrical protein. When I play around with the domain=20
analysis using DomainFinder I ususlly find one or more sets of modes,=20
deformation threshold and coarseness which return most of a single subuni=
t as=20
a domain, but only one of them. I'd assume that, at least in a totally=20
symmetrical mode  the symmetry equivalent atoms (or boxes) have to have=20
displacements that do not destroy the symmetry. Why do they NOT end up in=
=20
similar, symmetry related clusters? Either I don't get the argument given=
 in=20
the 1999 proteins article or something is strange. I understand that if t=
he=20
boxes are chosen without considering symmetry, slight violations of symme=
try=20
are to be expected. But not to see any symmetry at all?=20
Maybe someone can explain this to me?=20
Thanks a lot,
=09margot