[gmx-users] Unable to maintain proper density of polyunsaturated fatty acid

[Justin Lemkul]

On 3/5/15 2:27 PM, Tushar Ranjan Moharana wrote:
> Hi everyone,
> I am trying to make a bi-phasic system of water and triglyceride of poly
> unsaturated fatty acid. While making a box full of triglyceride of poly
> unsaturated fatty acid I encounter following problem.
> insert-molecule placed 126 molecule in 10nm *10nm*10nm cube. To get
> expected density around 500 molecule should be placed. I tried to do energy
> minimization followed by solvate the box with triglyceride but it did not
> work.
> During energy minimization of the above system E pot. remains positive
> (order of 10^4) and in later during *nVT* simulation all the molecule goes
> out of the box. When *nPT* simulation was done at 2 bar an error message
> come saying some molecules moved during simulation.
> I am using CHARM36 all atom force field and topology (.itp) file was
> created by swiss param. I simulated a single molecule in a box of
> 3nm*3nm*3nm. Everything went fine even though E pot after energy
> minimization was positive (order of 10^2) in this case too.
>
> Any suggestion to try is most welcome.
>

I would not expect gmx instert-molecules to succeed in building a well packed 
system in this case.  Inserting molecules as rigid entities randomly in a box 
works well enough for small molecules, but for something like a triglyceride, 
you have a large number of rotational degrees of freedom that will require 
extensive equilibration.  Build the box incrementally - fill some small box with 
as many TG as will fit, minimize and equilibrate for a long time.  Then increase 
the box size and add some more molecules, on and on until you reach whatever box 
size you need.

Positive energy may or may not mean anything.  In general, a condensed phase 
system should have a net negative energy based on large number of favorable 
nonbonded interactions.  But if the nonbonded interactions are weak relative to 
the bonded interactions, Epot is positive.  This is the case, e.g. for 1-octanol.

-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

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