Dear Chia-yun Chang,
Thanks for continuing to work through the tutorial and for asking
useful questions. :-)
> (1) How did I determine which atoms were relevant to the three dummy atoms?
The three dummy atoms (DM1, DM2 and DM3) are automatically created by
ProtoMS and are placed at the geometric centre of the molecule. They
are not shown in the diagram, which I agree is a little confusing. The
full diagram would be;
H06 DM1 H02
\ DM2 /
\ DM3 /
H07----C05------C01----H04
/ \
/ \
H08 H03
As the C01 atom is closest to the centre of the molecule, and thus
closest to the dummy atoms, the C01 atom is the one that is built
first and is the one that is bonded to the dummy atoms.
It doesn't really matter where the dummy atoms are, or which atoms
they are bonded to, as the dummies are used only to provide a frame of
reference for the coordinate system used by the z-matrix. The three
dummy atoms are built automatically to lie in a right-angle, e.g. the
bond DM1-DM2 is 90 degrees to the bond DM2-DM3.
> (2) The atom H02 and H03 seemed to have the same relative position.
> Why did the the topology lines have the differences between "H02 C01
> C05 H06" and "H03 C01 C05 H02"?
The z-matrix defines both the topology *and* how the atoms in the
molecule move. H02 and H03 do have the same topology, but are built in
the z-matrix to move in different ways. H02 is free to rotate around
the C01-C05 dihedral, while H03 is defined using an improper torsion,
so that it is constrained to lie in the same plane as C01-C05-H02.
This means that, as H02 rotates around the dihedral, so H03 rotates
around with it. This is a common pattern when building z-matricies for
Monte Carlo simulations, when you know that the movement of one atom
is closely linked to the movement of others, e.g. the three atoms of a
methyl (CH3) group should all rotate together, so we would build;
H3
C1-C2-H4
H5
H3-C2-C1-X
H4-C2-C1-H3
H5-C2-C1-H3
This would ensure that, as H3 rotated around the C1-C2 dihedral, so H4
and H5 would rotate around with it.
I hope this is helpful,
Best wishes,
Christopher