Ross,
That is good to know. Hopefully none of my suggestions (which are born
mostly out of necessary habits from my CHARMM days) would make things worse.
Last I looked at the code (or maybe this is just our local development
version), I think the coordinates are actually shifted by the QM region
center of geometry (not that I can imagine that making much of a difference
except for maybe really big QM regions). In any event, I am still confused
by how the upper/lower values in the sander bomb message are able to differ
by their sign when visualization of the trajectory appears to show the QM
atoms quite close together (maybe I used some really misleading
re-wrapping?). I'm not experiencing my old problems any more, but I still
don't know what exactly caused them other than, apparently, triclinic
boundary conditions.
Regards,
Brian
On Thu, Oct 6, 2011 at 12:18 PM, Ross Walker <ross.rosswalker.co.uk> wrote:
> Hi Brian et al,
>
> > 1) To make a single molecule from 2 things that are not, I think the
> > easiest
> > thing to do is use the "bond" command in leap and make a new prmtop.
> > I'd
> > have to look it up in the manual to tell you exactly how to use it.
> > Leap
> > may complain if any resulting bond, angle, or dihedral terms are not
> > defined
> > in your parm.dat file (hence the need for an additional frcmod). Keep
> > in
> > mind also that this might also affect 1-4 scaling of electrostatics,
> > but I
> > would assume that your QM region is large enough that this is not a
> > problem.
>
> You should really NOT need to mess with any of this. The QM code knows
> about
> periodic boundaries and should image things correctly. What happens
> internally in the code is that the first QM atom is moved to the origin and
> then everything is imaged around this giving you a QM region within the
> central box. Then the atoms that are within the cutoff are included and
> this
> makes essentially an imaged QM region with a qmcut buffer of MM atoms
> around
> it and this is what is used in the direct space part of the QM calculation.
>
> If you are having to mess with putting QM residues in the same molecule etc
> then there is something very wrong with your QM region.
>
> I suspect you are trying to run a QM region where the residues you are
> including are a long way away from each other or drift apart from each
> other
> during the simulation. The QM cut, for stability within the SCF calculation
> is a different to the traditional MM cutoff. Here ANY MM atom that is
> within
> QM cut of ANY QM atom is included in the list for ALL QM atoms. This means
> you essentially have a surface layer of thickness qmcut surrounding your QM
> region. Thus if you have two QM regions a long way apart you will get a
> bunch of MM atoms on one side of your box included in the direct space part
> of the QM region on the other side. When the code checks for the minimum
> image convention you run into problems because your QM-MM region is huge
> and
> overlaps the box edges.
>
> The code is simply not designed at present to deal with this situation.
> There are two approaches
>
> 1) Code in support for multiple QM regions. This needs careful thinking
> about to be able to do it in a formerly correct way, conserving energy etc.
>
> 2) Use an adaptive QMMM method to transform one of your QM molecules into
> MM
> when it diffuses away from the other or a given location.
>
> The second one is work in progress, the 1st is on the wish list.
>
> All the best
> Ross
>
> /\
> \/
> |\oss Walker
>
> ---------------------------------------------------------
> | Assistant Research Professor |
> | San Diego Supercomputer Center |
> | Adjunct Assistant Professor |
> | Dept. of Chemistry and Biochemistry |
> | University of California San Diego |
> | NVIDIA Fellow |
> | http://www.rosswalker.co.uk | http://www.wmd-lab.org/ |
> | Tel: +1 858 822 0854 | EMail:- ross.rosswalker.co.uk |
> ---------------------------------------------------------
>
> Note: Electronic Mail is not secure, has no guarantee of delivery, may not
> be read every day, and should not be used for urgent or sensitive issues.
>
>
>
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
================================ Current Address =======================
Brian Radak : BioMaPS
Institute for Quantitative Biology
PhD candidate - York Research Group : Rutgers, The State
University of New Jersey
University of Minnesota - Twin Cities : Wright-Rieman Hall 101
Graduate Program in Chemical Physics : 610 Taylor Road,
Department of Chemistry : Piscataway, NJ
08854-8066
radak004.umn.edu :
radakb.biomaps.rutgers.edu
====================================================================
Sorry for the multiple e-mail addresses, just use the institute appropriate
address.
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Oct 06 2011 - 10:00:03 PDT
