[gmx-users] virtual site - algorithmic reasons

[Justin Lemkul]

On 3/31/16 6:06 AM, Samith Rathnayake wrote:
> Hi everyone,
>
> Recently i was used to started Bevan Lab tutorial  and gone through
> specially with virtual site tutorial.In this tutorial Dr.Justin emphasized
> that co2 cannot effectively build in traditional sense  because of
>   algorithmic reasons. And cant hold 180 degree during a simulation.Could
> any one elaborate what are this "algorithmic reasons" ? any publication
> upon this virtual site techniques in any gromos force field? and also about
> this algorithmic reasons ? And if it is okey to used this virtual site
> techniques in gromacs 4.6.5 version ?
>

There are mailing list posts about it; I forget the details but the long and 
short of it was that the fluctuations around a 180-degree angle lead to 
numerical instability in the code.  IIRC that was fixed some time ago, but I 
would not expect 4.6.5 to work.  Try the newest version and see if you can get 
away with a linear angle without any fancy work.

-Justin

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

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

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