Re: Angle of lone pair in TIP4P 2005 simulation

From: Peter Freddolino (petefred_at_umich.edu)
Date: Tue May 08 2018 - 23:06:55 CDT

On Tue, May 8, 2018 at 8:57 PM, Vermaas, Joshua <Joshua.Vermaas_at_nrel.gov>
wrote:

> Did you ever call "regenerate angles dihedrals" in building your psf? Or
> used something else that automatically generates angles if two atoms share
> a neighboring atom? NAMD is clearly telling you that you've built an angle
> there one way or another when you prepared the psf. The obvious solution,
> if you are positively sure that there shouldn't be an angle term there, is
> to just add a angle into your parameter set with a zero force constant. A
> better solution would be to figure out where in your psf construction
> process these angles are being introduced, and eliminate them.
>

On the particular point of this angle -- because NAMD doesn't apply
constraints like SETTLE during minimization, it is actually useful both to
have that angle there *and* to have a nonzero force constant, to keep the
LP geometry reasonable during minimization. In the TIP4 files I had used
for validating the implementation, I had the entry
HT OT LP 50.0 52.26
to define it. Note that once dynamics start, because the water is
constrained to its ideal geometry anyway, by definition this angle term
will have zero effect.
Best,
Peter

> On 2018-05-08 18:21:47-06:00 owner-namd-l_at_ks.uiuc.edu wrote:
>
> Hi George,
> You described the situation very well!
> Best,
> Esmael
>
>
> On May 8, 2018, at 7:08 PM, Geo Reuter <georeuter_at_gmail.com> wrote:
>
> Hello,
> I am trying to do a minimization of a TIP4P 2005 waterbox and got this
> error:
> FATAL ERROR: UNABLE TO FIND ANGLE PARAMETERS FOR HT OT LP (ATOMS 3 1 2)
> This confuses me because from my understanding there should be no angle
> for the lone pair in the TIP4P 2005 water model. The topology file I used
> contains no angle so I thought this was the case. Does anyone have
> experience using TIP4P 2005 and could tell me what went wrong in my
> minimization.
>
>

This archive was generated by hypermail 2.1.6 : Tue Sep 17 2019 - 23:19:52 CDT