-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hello Tom--
>
> I am trying to do a simulated annealing with RDC and NOE restraints
> but with explicit electrostatics turned on. The following param
> statement seems to work ok:
>
> parameters
> nbonds
> atom
> nbxmod 4 {* This is set to 4 (bonds), because we are
> using *}
> {* internal dynamics (torsion angle
> refinement). *}
> wmin = 0.01 {* warning
> off *}
> tolerance=0.5
> eps=30 cutnb=11.5 ctonnb=9.5 ctofnb=10.5 tolerance=0.5 rdie
> vswitch switch
> end
> end
> (Yes, I stole that from refine_gentle.inp)
>
> And the electrostatics are turned on during the initial minimization
> and through the first steps torsional high temperature dynamics.
> However, once the cool loop starts (itype=pc6) and vanderWaals radii
> start increasing, if I use the repel=$radius as in:
>
> parameter
> nbonds
> cutnb=4.5 rcon=$k_vdw nbxmod=4 repel=$radius
> end end
>
>
> This appears to turn off the elec term in the energy function.
> However, getting rid of repel results in huge VanderWaals energies
> and distorted structures. Is there a way around this?
>
There were problems with the RDC pseudo atoms and VDW for non-repel
potentials in older versions of Xplor-NIH. Are you using 2.25?
Anyway, I do finally seem to have a setup for water refinement which
seems to work. If you're interested, I will post the scripts.
Charles
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Processed by Mailcrypt 3.5.9 <http://mailcrypt.sourceforge.net/>
iEYEARECAAYFAkwRE6EACgkQPK2zrJwS/lYlRwCeLUdcLibA2BhDxmHj5kV9E9Yt
XGUAniJ5BMbm9RQ/DImPqqkEOla0fWYd
=VTje
-----END PGP SIGNATURE-----
