[gmx-users] The pullf.xvg and pullx.xvg files

[Laura Tociu]

Dear Justin,

I have done two parallel short pullings, one with pull_geometry = cylinder
and one with pull_geometry = direction.

When pull_geometry = direction, everything works perfectly: the grompp
output says that the distance at start, as well as the reference at t=0 are
both equal to 3.019 nm, and the pullx.xvg file reflects this: at t=0 the
distance is also 3.019 nm. This distance is reasonable based on my
geometry. The force at t=0 is 10^-6 and the COM pull energy in the md.log
file is on the order of 10^-10.

When pull_geometry = cylinder, the grompp output says the distance at start
and the reference distance are 1.215 nm (which they aren't) but the
distance at t=0 in the pullx.xvg file shows up as 2.8 nm (which it is
roughly -  it's hard to tell where the COM as computed by the cylinder
approach would be but it is likely to be close to the COM as usually
computed). This huge discrepancy leads to huge forces as well as energies
at t=0.

Is this some kind of bug that has been fixed since Gromacs 5.0.2, which is
the version I'm using, or should I report it?  I think since the distance
computed by grompp and that in the pullx.xvg file at t=0 are different, the
method will likely give inaccurate forces and thus PMF's... What do you
think?

Laura



On Thu, Jun 25, 2015 at 11:07 AM, Laura Tociu <ltociu at princeton.edu> wrote:

> Ok thanks! I will analyze this more deeply and maybe also try differen
> geometries. And sorry for the stupid question, I was reading force but was
> thinking potential instead...
>
> Laura
> ------------------------------
> From: Justin Lemkul <jalemkul at vt.edu>
> Sent: ‎25/‎06/‎2015 02:23
> To: gmx-users at gromacs.org
> Subject: Re: [gmx-users] The pullf.xvg and pullx.xvg files
>
>
>
> On 6/24/15 9:20 AM, Laura Tociu wrote:
> > Dear Justin,
> >
> > Thanks for the reply! Yeah, I understand how the pulling works now. The
> > forces at time t=0 are not zero, though. There are huge forces such as
> -270
> > or even -1700 kj/mol/nm acting on my pull group at time t=0.  What do you
> > believe could be the cause of that? And why is there a /nm in that
> force? I
> > mean, isn't the force per nm (force constant) of the spring always 1000
> > kj/mol/nm, but the actual force adopts various different values as time
> > goes by?
> >
>
> The force constant is in kJ/mol/nm^2 (same as all bonded force constants).
> Force is kJ/mol/nm because force is the negative derivative of potential
> with
> respect to position.  You can also convert this rather easily to pN or
> some more
> familiar unit.
>
> > I ran a short simulation with pull_coord1_rate = 0, and when I did that I
> > got reasonable forces such as 40-60 kj/mol/nm (??) but still not a zero
> > force at time t=0. Please let me know if this is normal behavior or not.
>
> Sorry, can't tell without fully analyzing all of your files (not something
> I
> have time for).  Check what grompp reports as the reference distance at
> t=0 vs.
> what you calculate in the input coordinate file/first frame of the
> trajectory.
> I've never used the cylinder geometry; could be something specific to that.
>
> -Justin
>
> --
> ==================================================
>
> Justin A. Lemkul, Ph.D.
> Ruth L. Kirschstein NRSA Postdoctoral Fellow
>
> Department of Pharmaceutical Sciences
> School of Pharmacy
> Health Sciences Facility II, Room 629
> University of Maryland, Baltimore
> 20 Penn St.
> Baltimore, MD 21201
>
> jalemkul at outerbanks.umaryland.edu | (410) 706-7441
> http://mackerell.umaryland.edu/~jalemkul
>
> ==================================================
> --
> 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.
>