Re: [AMBER] cpptraj gist: Eww energies

From: Thomas Fox <>
Date: Wed, 12 Aug 2015 17:31:09 +0200

   Thanks to all who responded to my question...yes, RMS fit as the culprit
   makes perfectly sense...I played around with
   constraints, and one of the attempts was restraining just one of the
   5membered rings: without RMSD fit, things look good, i.e. no voxel with Eww
> 0, with RMSD fit, I get a symmetrical pattern of high energy voxels at the
   faces of the gist box. Looking at the picture is very convincing that the
   rotation is the cause of the problem...

   One reason I decided not to apply restraints (as suggested in the paper and
   in the tutorial) was that I wanted to see movements of the solute -
   otherwise I probably could have used 3D-RISM at a much lower cost...but
   Tom's suggestion how to pin down the solute without losing its dynamics
   sounds good.

   Thanks again,
   Gesendet: Montag, 10. August 2015 um 19:54 Uhr
   Von: "Tom Kurtzman" <>
   An: "AMBER Mailing List" <>
   Betreff: Re: [AMBER] cpptraj gist: Eww energies
   Yeah, Crystal wrote all the initial code for GIST and knows the
   implementation better than anybody.
   rereading the emails from the other person with this problem (which Crystal
   also solved), there was both a problem with imaging (not using familiar
   with an octahedral box) but also with the RMS fit. The RMS fit does seem
   consistent with the problems you are having. If you cut out the voxels
   within a a few angstroms (3.5+ about), where the minimum image makes no
   difference then you should be okay with the analysis you have.
   That being said, with GIST, as Dan and Crystal mention, it's probably best
   just to restrain enough heavy atoms in the system to prevent the
   translation and rotation of the solute altogether. For proteins, a
   technique which we often use is to restrain a few sites (at least 3 non
   linear heavy atoms) distal from the region of the interest (often a protein
   active site) and preferably restrain heavy atoms with low b-factors. The
   aim of this is to maintain as much important motion in the system while
   simultaneously preventing the translation and solute. We also tend to run
   systems with all heavy atoms restrained and with main chain heavy atoms
   restrained. We generally run all three (each is informative in its own
   On Mon, Aug 10, 2015 at 8:12 PM, Daniel Roe <> wrote:
> Hi,
> On Mon, Aug 10, 2015 at 3:49 AM, Thomas Fox <> wrote:
> > Eww values equal 0, but then 8 voxels with energies up to 300
> kcal/mol. The
> > neigboring voxels of these high-energy ones have what I would consider
> > "normal" values.
> >
> > This behavior becomes extremely pronounced when I use a larger grid
> for my
> > gist calculation (for a 41x41x41 grid with a spacing of 0.5A, I get
> energy
> > values up to 1E19, and about 13000 of the 68921 data points have a
> Eww-norm
> > > 100 kcal/mol.
> >
> >
> > parm
> > trajin cb7_solv.crd
> > trajin 1 5000
> > autoimage origin
> > rmsd CUC :1 first
> I would bet the large values you are seeing are the result of
> collisions caused by RMS-fitting after imaging (as mentioned by
> Crystal). The problem is that GIST uses imaged distances in it's
> calculations, but once you RMS-fit a system it is almost always
> rotated away from the original unit cell axes and so imaging no longer
> works properly (which is why cpptraj prints out a warning to this
> effect when RMS-fitting a system with PBC). For small rotations the
> effect may not be noticeable, but it's there. The effect is more
> pronounced for atoms farther away from the rotation which is why
> increasing the grid size gives you more clashes.
> As implied by Crystal probably the best way to ensure you get a well
> behaved grid at the moment is to put some mild positional restraints
> on your solute. On my current to-do list is to have a grid which can
> "move" with the solute using the RMS results - I hope to have this
> implemented soon.
> Hope this helps,
> -Dan
> >
> > gist gridcntr 0. 0. 0. griddim 21 21 21 gridspacn 0.5 out
> gist_21_05.out
> >
> > run
> > _______________________________________________
> > AMBER mailing list
> >
> > [1]
> --
> -------------------------
> Daniel R. Roe, PhD
> Department of Medicinal Chemistry
> University of Utah
> 30 South 2000 East, Room 307
> Salt Lake City, UT 84112-5820
> [2]
> (801) 587-9652
> (801) 585-6208 (Fax)
> _______________________________________________
> AMBER mailing list
> [3]
   Tom Kurtzman, Ph.D.
   Assistant Professor
   Department of Chemistry
   Lehman College, CUNY
   250 Bedford Park Blvd. West
   Bronx, New York 10468
   AMBER mailing list


AMBER mailing list
Received on Wed Aug 12 2015 - 09:00:03 PDT
Custom Search