From: Bill Ross <ross.cgl.ucsf.EDU>
Date: Sat, 26 Feb 2011 12:17:51 -0800

> Are you using SHAKE for the water? Is there the "extra" H-H bond in water
> that the TIP3P parameters assume? Both of these are required if you use the
> parameters for TIP3P water. (It's the Lennard-Jones parameters of the oxygen
> that keeps the hydrogens from merging with nearby molecules.

I.e. SHAKE keeping bond lengths fixed guarantees the H's remain
within the O's vdw. The H-H bond is an easy way to constrain the
HOH angle.

It sounds as if you defined the atom type's hybridization in leap?
And that after this the pdb from inpcrd+prmtop->ambpdb became planar
in the ring(s) of interest. The only way I can imagine the definition
of hybridization to matter to the leap-generated inpcrd is if the
input pdb lacked coordinates and leap needed to generate them using
rules or guesswork it seldom needs to rely on.

Had your bad structures been the result of leap->inpcrd+prmtop->sander,
then I would worry that defining hybridization type might need to be
moved to the parm.dat and frcmod files, since it could affect the
way impropers are defined via the generation of lists of atoms
that might be centers for improper dihedral parameter application.


AMBER mailing list
Received on Sat Feb 26 2011 - 12:30:06 PST
Custom Search