Re: [AMBER] Computing forces by trajectory post-processing with sander

From: Daniel Roe <>
Date: Tue, 23 Apr 2019 09:56:30 -0400


There was a bug in the force write for sander imin=5 post processing
that is fixed in the upcoming AmberTools 19 release (soon). If you
need it immediately I can try to guide you how to manually patch the


On Mon, Apr 22, 2019 at 3:54 PM Casalini Tommaso
<> wrote:
> Dear Amber users and developers,
> I would like to compute the forces acting on a certain number of atoms in my molecular trajectory.
> In order to do this, I used sander with the following input file:
> Computing forces
> &cntrl
> imin=5,
> irest=1,
> ntx=5,
> ntwf=1,
> ioutfm=0,
> ntwprt=2000,
> ntb=2,
> ntp=1
> /
> I run sander with the appropriate flags, indeed everything goes smoothly and the computed energies make sense (I compared them with the previously obtained values from the simulations). Anyway, the file "forces.mdcrd" where I wanted to store the computed forces is empty.
> I had another look at the manual, which states:
> "Every ntwf steps, the forces will be written to the mdfrc file. If ntwf = 0, no force trajectory file
> will be written. If ntwf = -1, forces will be written to the mdcrd, which then becomes a combined
> coordinate/force trajectory file, at the interval defined by ntwx. This option is available only for binary
> NetCDF output (ioutfm = 1)."
> At this point, it is not clear to me whether the ntwf-not-equal-to-zero option is compatible only with ioutfm = 1 in every case, or if this is strictly true only when ntwf = -1. Indeed, I obtained an empty file but sander did not complain.
> My questions are the following:
> 1) Why did I obtain an empty file for the forces? Did I make some mistake, or should I put ioutfm = 1?
> 2) Assuming that I must set ioutfm = 1, I would like to have an ASCII formatted file for the force, since I wrote a post-processing code in matlab that can be "fed" with .mdcrd files. Can I make a Netcdf-to-ascii conversion with cpptraj as I do for molecular trajectories?
> I thank you in advance for your help and support.
> With my best regards,
> Tommaso
> _______________________________________________
> AMBER mailing list

AMBER mailing list
Received on Tue Apr 23 2019 - 07:00:03 PDT
Custom Search