[gmx-users] Parrinello-Rahman and Pairtype1 causing pressure instabilities

[Wesley Michaels]

Hey y'all,

Ran some additional simulations and, unfortunately, I'm still experiencing
problems. Here's the mdp file that I used:

integrator              =  md
nsteps                  =  20000000 ; 20 ns
dt                      =  0.001000
nstlist                 =  20
ns_type                 =  grid

nstxout                 =  25000
nstvout                 =  25000
nstenergy               =  25000
nstlog                  =  25000

continuation            =  yes
constraints             =  none
lincs_iter              =  1
lincs_order             =  4
cutoff-scheme           =  Verlet

rcoulomb                =  1.4
rvdw                    =  1.4
coulombtype             =  PME
vdwtype                 =  PME
pme_order               =  4
fourierspacing          =  0.12

tcoupl                  =  Nose-Hoover
tc-grps                 =  System
tau_t                   =  0.100000
ref_t                   =  300.000000
nsttcouple              =  5

pcoupl                  =  Parrinello-Rahman
pcoupltype              =  isotropic
tau_p                   =  1.000000
ref_p                   =  1.000000
compressibility         =  0.000045

refcoord_scaling        =  com
pbc                     =  xyz
periodic-molecules      =  no
DispCorr                =  EnerPres
gen_vel                 =  no


The simulation ran for 40ns with the Berendsen barostat just fine - density
and volume were well-equilibrated within 1ns of the simulation starting.
Unstable density fluctuations began immediately after switching to the PR
barostat (from the output configuration of the previous 40ns run) and the
simulation failed within 0.4ns.

Here are some links to:
- topology file:
https://drive.google.com/open?id=1NF5EPbM2mnTGilXpOf7pQK5qlelBd8KC
- initial configuration:
https://drive.google.com/file/d/1kxMOKQcsAEheBCEAbkGcXqfBDwsHKpwr/view?usp=sharing
- log file of second half of Berendsen run:
https://drive.google.com/file/d/15TgbOOr6_fb8FkqR-nFHR8j1qbhTNT-Q/view?usp=sharing
- log file of PR run:
https://drive.google.com/file/d/1CPobb3ddDsd4I6Fce1tEpwHiBd5yq-tE/view?usp=sharing

Thanks! If you need any other files or information, please let me know.

-Wes




On Tue, Mar 24, 2020 at 3:44 AM Justin Lemkul <jalemkul at vt.edu> wrote:

>
>
> On 3/24/20 3:09 AM, Wesley Michaels wrote:
> > Hi,
> >
> > Thanks for your reply! I built the topology with my own script, as
> opposed
> > to using pdb2gmx. Following Bogdan's instructions here (
> >
> https://mailman-1.sys.kth.se/pipermail/gromacs.org_gmx-users/2012-December/077005.html
> ),
> > I specified all instances of 1-4 pair interactions (i.e. pairs of atom
> > indices) in the [ pairs ] section, but let GMX specify the [ pairtypes ]
> > section (i.e. the interaction parameters for pairs of atom types)
> > automatically using "gen_pairs" = "yes".
> >
> > As a second test, I generated a topology using pdb2gmx and the procedure
> > Justin detailed here (
> >
> https://mailman-1.sys.kth.se/pipermail/gromacs.org_gmx-users/2009-March/040125.html
> )
> > (using the OPLS-AA FF option in pdb2gmx) and used that in a trial
> > simulation. I experienced the same issue, unfortunately. Happy to provide
> > any files as needed to help clarify the issue.
>
> If pdb2gmx built the topology and you're using an unaltered
> forcefield.itp [defaults] directive, based on what you've said below, it
> appears to just be an equilibration issue. Use Berendsen for a while to
> get the system settled and then switch to Parrinello-Rahman. If you're
> still experiencing problems, full .mdp file(s) would be helpful.
>
> -Justin
>
> --
> ==================================================
>
> Justin A. Lemkul, Ph.D.
> Assistant Professor
> Office: 301 Fralin Hall
> Lab: 303 Engel Hall
>
> Virginia Tech Department of Biochemistry
> 340 West Campus Dr.
> Blacksburg, VA 24061
>
> jalemkul at vt.edu | (540) 231-3129
> http://www.thelemkullab.com
>
> ==================================================
>
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>