[gmx-users] Re: periodic molecule

[Justin Lemkul]

On 9/18/13 8:07 AM, Valentina wrote:
> Thank you!
>
> Yep, they seem to be wanting to bond to the O's within the cell, while they
> are meant to bond across the PBC.
>
> I am not sure how periodic_molecule = yes part functions. I don't seem to
> get difference if I turn it on or off when I do Energy minimisation. Do you
> have some examples that work, so I can compare to them? Are there some

I have never used periodic molecules, except when toying around with some CNTs 
for my own curiosity.  I suspect the problem you're seeing arises from the 
bonded atoms being closer within the unit cell than they are across PBC, so the 
minimum distance is what is trying to be satisfied.  How far apart are the atoms 
across PBC, and how long is the assigned bond supposed to be?  I have seen other 
cases posted to the list wherein the periodic_molecules function breaks down 
simply because of bad starting geometry.

> threshold distances, that they rely upon? I really wouldn't like to have to
> make the a bigger basic cell - too much topology assigning.
>

A bigger unit cell won't help.  In fact, it would probably make things worse.

-Justin

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

Justin A. Lemkul, Ph.D.
Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441

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