The leaves of the tree don't have to be atoms. So one protein could
have atoms with xyz as leaves and another domains with xyz and
orientation as leaves (and it could not have atoms as leaves).
On May 24, 2009, at 2:13 PM, Friedrich Foerster
On Fri, May 22, 2009 at 2:19 AM, Daniel Russel <>
Slight improvement of the proposal:
- if any of the hierarchy has coordinates, all the leaves must
just one q: what about systems with different granularities; i.e., for
part of a complex one might have (in fact rather often...) reasonable
atomic models for some proteins or domains and not for others. for
example, i sometimes have fragments, which have coordinates, and for
others i can go finer in the tree.
how is that solved?
- one must be able to determine which Particle each residue is
- the default "radius" attribute is used to an approximating
radius- so the
radius that best approximates the sub-particles at that level of
- hierarchy particles can have a "bounding radius" attribute which
represents a bounding sphere. This will be used by the close pairs
other code and is assumed to include all relevant geometry
the particle (so all geometry in the subtree).
On Thu, May 21, 2009 at 10:07 AM, Daniel Russel <>
I suggest that we impose that in an atom.Hierarchy
- if any of the hierarchy has coordinates, all the leaves do (they
at any level of the hierarchy though)
- all residues in each molecule must be either a) explicitly
as a Residue, b) part of a Domain which is a leaf (which would be
to allow non-contiguous sets of residues--or we could split off a
of decorator for this)
- any member of the hierarchy other than an Atom which has
also has a radius. The interpretation is that this radius bounds
members of the (possibly elided) subtree can be.
These would ensure several useful properties:
1) it be clear what the highest resolution representation of any
2) one be able to find out where in the hierarchy any residue is
3) code does not need to look at the most detailed representation
to get a
rough picture of the structure, it only has to search down the
from CS) to find a representation with small enough balls.