[gmx-users] Positive potential energies for Dimethyl Aniline and Aniline

[Justin A. Lemkul]

On 5/6/12 11:42 AM, Debayan Chakraborty wrote:
> Dear Colleagues,
>                            I am trying to build a well equilibrated solvent box
> of aniline and N,N dimethyl aniline using the OPLS-AA parameters. I have built
> my topolgy files by closely following that given for 2-chloroaniline in the
> molecule and liquids database of virualchemistry.org
> <http://virualchemistry.org>. I have even downloaded the files of chloroaniline
> and ran short simulations. In those simulations I get negative potential and
> total energies, and even the Coulomb-recip term is negative.
>
> When I switch to my systems aniline and DMA, for both I get negative potential
> energies for the minimized structure. However, the coulomb-recip term is
> positive in both cases. When I start my MD total energies for both aniline and
> DMA become positive, though aniline still retains a negative value of potential
> energy, the potential for DMA becomes positive.
>
> My questions are : Are positive or negative energies of a system during
> simulation real indicators of something wrong in your MD ?

A positive potential indicates net repulsion.  In a normal, condensed-phase 
system, there should be net attraction.

>                              How is it possible that chloroaniline, aniline and
> DMA which are closely related systems have so different values of energies ?

"Closely related" is in the eye of the beholder.  Even small changes to chemical 
substituents can produce wildly different properties.  Yes, the molecules have 
the same general structure and "shape," but modifications of chemical groups can 
change the very nature of the molecule.  For instance, dimethylation of the 
amine on aniline removes its ability to form hydrogen bonds.

>                              Does a positive Coulomb recip term indicate wrong
> representation of electrostatics ?
>

Again this indicates net repulsion in the long-range electrostatic term (from 
PME).  It probably indicates that your model is not adequate, as the molecules 
are likely to be repelled from one another.

> For all the systems I have used the charges from the OPLS-AA force field. Any

How so?  If you're deriving parameters for a new topology, what exactly are you 
using?  Charges are not necessarily transferrable between molecules, though many 
common groups are.  Be careful what you interpret as "similar enough" if this is 
what you're doing.

-Justin

-- 
========================================

Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin

========================================