Re: problem with water molecules angle in MD simulation

From: Axel Kohlmeyer (akohlmey_at_gmail.com)
Date: Fri Jan 30 2015 - 05:32:06 CST

On Fri, Jan 30, 2015 at 3:43 AM, Tristan Croll <tristan.croll_at_qut.edu.au> wrote:
> That's a question for someone who knows the guts of NAMD a lot better than

this is not a NAMD thing, but a general MD problem: constraints via
SHAKE only work through MD, not minimization. There is no equivalent
solver algorithm. Some MD codes (silently) replace the constraints
with a very stiff harmonic potential. For codes that don't do this,
one can modify the force constant parameters for the water bond/angle
correspondingly. it won't affect MD, since the force constant is
ignored with SHAKE.

axel

> I. For the time being, if you're worried about it a simple workaround would
> be to write an extrabonds file to maintain either the angle or the H-H
> distance during minimization. Try this in the tkconsole:
>
>
> set outfile [open extrabonds-water.txt w]
>
> set bondstrength 200
>
> set Hdist 1.598
>
> set water [atomselect top water]
>
> foreach res [lsort -unique [\$water get residue]] {
>
> set thisH1 [atomselect top "residue \$res and name H1"]
>
> set thisH2 [atomselect top "residue \$res and name H2"]
>
> set H1index [\$thisH1 get index]
>
> set H2index [\$thisH2 get index]
>
> puts \$outfile "bond \$H1index \$H2index \$bondstrength \$Hdist"
>
> }
>
> close \$outfile
>
>
> Then run your minimization with this extrabonds file applied, and remove it
> for the equilibrium dynamics.
>
>
> ________________________________
> From: Shoresh Shafei <s.shafei_at_gmail.com>
> Sent: Friday, 30 January 2015 6:28 PM
> To: Tristan Croll
> Cc: namd-l_at_ks.uiuc.edu; Norman Geist
> Subject: Re: namd-l: problem with water molecules angle in MD simulation
>
> Tristan,
>
> Thank you for tying it on your machine. The questions is why this is
> happening. Did you have any explanation for it?
>
> Best Regards
> Shoresh
>
> On Jan 30, 2015, at 3:25 AM, Tristan Croll <tristan.croll_at_qut.edu.au> wrote:
>
> I tried this out on my machine using TIP3P water, the CHARMM-36 forcefield,
> and NAMD 2.10, and I see the same thing. During energy minimization the
> water molecules adopt a wide range of angles down to about 95 degrees, but
> they all snap back to ideality as soon as the equilibrium simulation starts.
>
>
> Tristan
>
> ________________________________
> From: owner-namd-l_at_ks.uiuc.edu <owner-namd-l_at_ks.uiuc.edu> on behalf of
> Norman Geist <norman.geist_at_uni-greifswald.de>
> Sent: Thursday, 29 January 2015 5:56 PM
> To: 'Shoresh Shafei'
> Cc: namd-l_at_ks.uiuc.edu
> Subject: AW: namd-l: problem with water molecules angle in MD simulation
>
>
> Couldnâ€™t see anything unsual. Have you already tried recreating the .top
> file, sometimes there are corruptions which lead to such behavior?
>
>
>
> Norman Geist.
>
>
>
> Von: Shoresh Shafei [mailto:s.shafei_at_gmail.com]
> Gesendet: Mittwoch, 28. Januar 2015 17:21
> An: Norman Geist
> Betreff: Re: namd-l: problem with water molecules angle in MD simulation
>
>
>
> Hi Norman,
>
> The minimization and constant T simulation input files are attached.
>
> The rigid bond option did not work when used in Minimization simulation.
>
> Thank You!
>
> Shoresh
>
>

```--
Dr. Axel Kohlmeyer  akohlmey_at_gmail.com  http://goo.gl/1wk0
College of Science & Technology, Temple University, Philadelphia PA, USA
International Centre for Theoretical Physics, Trieste. Italy.
```

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2015 - 23:21:36 CST