Re: unreasonably high temp/pressure. .

From: Jesper Sørensen (jesorensen_at_ucsd.edu)
Date: Wed Oct 24 2012 - 23:24:42 CDT

Hi Erik,

I am successfully using the TIP4P-ew water molecules in the Amber99SB force field running with NAMD 2.8 and 2.9, so it should work.
I'll look through and compare your conf files with mine sometime tomorrow and see what differences I can spot.

Do you have a tleap script that I can see, just to check the initial parm generation?

Best,
Jesper

On Oct 24, 2012, at 3:18 PM, "Martin, Erik W" <Erik.Martin_at_stjude.org> wrote:

> I honestly don't know if TIP4P is meant to be be rigid or not… but perhaps it does need to be run with rigid water? If that’s the case, its a little surprising that no mention of this is made in any of the documentation I found online. Leaving "langevinHydrogen" off is, oddly some artifact of carrying over and modifying old configuration files. I obviously set it off years ago for some reason, but can't for the life of me figure out what that was.
>
> I created my structure in leap, I used the same starting peptide pdb, but had leap add the tip4 box. I have actually extensively minimized this system. I can see how looking at the conf files I attached don't make it seem that way. The first one I attached used as a starting structure a minimized trajectory and then had an additional 1000 min steps. As for the temperature ramping input I last attached, I think my current file only had like 50 min steps so that I could get past them quickly to see if the actual simulation was running properly (which it obviously was not). I think I originally had this set at something like 10k minimization steps. I checked the energies, and they seemed to have gotten to a steady state.
>
> Thanks a lot for the help. Its good to hear that I'm not the only one confused by this,
> Erik
>
> From: Aron Broom <broomsday_at_gmail.com<mailto:broomsday_at_gmail.com>>
> Date: Wed, 24 Oct 2012 17:04:54 -0500
> To: Jesper Sørensen <jesorensen_at_ucsd.edu<mailto:jesorensen_at_ucsd.edu>>
> Cc: "namd-l_at_ks.uiuc.edu<mailto:namd-l_at_ks.uiuc.edu>" <namd-l_at_ks.uiuc.edu<mailto:namd-l_at_ks.uiuc.edu>>
> Subject: Re: namd-l: unreasonably high temp/pressure. .
>
> I thought that TIP4P was parameterized as a rigid molecule? Anyway, related to this topic, I see you have "langevinHydrogen off", which means the thermostat doesn't apply itself to the hydrogens. In the case where TIP4P is rigid, this would be ok, but now those hydrogens aren't rigid, so maybe try "langevinHydrogen on" (comes with an ~5% performance penalty) or "rigidbonds water".
>
> Honestly I don't see how either of the above things I'm suggesting would solve a problem of this magnitude, but I think they are important anyway.
>
> So, how did you convert your TIP3P simulation to TIP4P? Did you take out the peptide and re-solvate with TIP4P? Or did you just rename the waters and toss in a forth point? If the latter is the case, then I can imagine the coordinates for the 4 points of the water might be starting fairly messed up, and you would want to energy minimize for AT LEAST 1 step per water molecule. Maybe I'm crazy on that, but this really seems like it might just be a poorly minimized system, no?
>
> ~Aron
>
> On Wed, Oct 24, 2012 at 4:00 PM, Jesper Sørensen <jesorensen_at_ucsd.edu<mailto:jesorensen_at_ucsd.edu>> wrote:
> Hi Aron,
>
> You probably did this but just in case:
>
> When you generated your prmtop did you set "set default FlexibleWater on" in LEaP?
> http://ambermd.org/namd/namd_amber.html, section 5
>
> Best,
> Jesper
>
>
> On Oct 24, 2012, at 8:26 AM, Aron Broom <broomsday_at_gmail.com<mailto:broomsday_at_gmail.com>> wrote:
>
> Did you tell NAMD you are using TIP4P? you have to put "waterModel tip4" in your config file, the default is tip3.
>
> ~Aron
>
> On Wed, Oct 24, 2012 at 10:55 AM, Martin, Erik W <Erik.Martin_at_stjude.org<mailto:Erik.Martin_at_stjude.org>> wrote:
>
> I'm hoping someone on this list will have some insight on this problem I've been having. I'm running a simulation of an 82aa peptide in a truncated octahedron water box for a total of ~100k atoms. I'm also using an AMBER ff. I have been running this simulation for ~25ns using tip3 water with no problems at all. However, I wanted to run this with tip4 water as well and am seemingly incapable of getting it to equilibrate properly.
>
> On the first structure, I was able to equilibrate it by constraining peptide coordinates and allowing the water – with rigid bonds – to equilibrate in the cell. When I rebuilt the system with tip4 water I did absolutely nothing different. It runs fine with rigid bonds on and the peptide constrained, but when I try to allow it to run I get the dreaded "atoms moving too fast" error at about step 120-200. It is different atoms every time, and what seems to be happening is that the temperature and pressure both ramp up to enormous values. (for the record, I'm using Langevin dynamics to control temp and pressure)
>
> Does anyone have a clue what could be causing this condition?
>
> Thanks a lot,
> Erik
>
> ________________________________
> Email Disclaimer: www.stjude.org/emaildisclaimer<http://www.stjude.org/emaildisclaimer>
> Consultation Disclaimer: www.stjude.org/consultationdisclaimer<http://www.stjude.org/consultationdisclaimer>
>
>
>
>
>
> --
> Aron Broom M.Sc
> PhD Student
> Department of Chemistry
> University of Waterloo
>
>
>
>
>
> --
> Aron Broom M.Sc
> PhD Student
> Department of Chemistry
> University of Waterloo
>
>

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:22:12 CST