[gmx-users] Nucleic Acid in vacuo

[Justin A. Lemkul]

Matthias Ernst wrote:
> Thank you very much for your help, Justin. 
> I tried it with those mdp options:
> 
> title           =       INITIAL_EM0
> cpp             =       /lib/cpp
> constraints     =       all-bonds
> integrator      =       md
> rlist           =       0.0
> rcoulomb        =       0.0
> rvdw            =       0.0
> pbc 		= 	no
> 
> This resulted in
> Angle    Proper Dih.  Improper Dih.          LJ-14     Coulomb-14
> 1.88961e+01    6.04179e+01    5.60480e-04    3.74069e+01   -2.06394e+02
> LJ (SR)   Coulomb (SR)      Potential    Kinetic En.   Total Energy
> 6.94305e-03   -5.66241e+01   -1.46290e+02    2.88457e-01   -1.46002e+02
> Temperature Pressure (bar)   Constr. rmsd
> 1.21730e+00    0.00000e+00    0.00000e+00
> 
> So, now the 1-4-values are slightly better (37.11 <-> 37.04 LJ, -207.7
> <-> -206.3 Coulomb), while the SR-LJ and Coulomb energys are at two
> orders of magnitude too small (LJ: 0.7127 <-> 0.0069430, Coulomb:
> -4354.5 <-> -56.62). 
> 
> In fact, I considered looking through the code but wanted to avoid it
> because I fear that there are so much optimisations implemented in
> Gromacs that I probably will not understand whats going on and why.
> Maybe I will try it, but first I will play with the forcefield
> parameters to see what happens if I switch everything off except certain
> pairs, then I can compare them.
> 
> Besides, I'm wondering, is there a way to disable all optimisations and
> just do a "plain" Lennard-Jones and Coulomb evaluation in vacuo without
> any box, solvent, boundary conditions and without cutoffs in Gromacs?
> 

That's essentially what you're doing with the .mdp file above.  No box (pbc = 
no), no cutoffs (r* = 0) and no solvent.  There are options to disable 
optimizations and use generic kernels in Gromacs, but I can't recall all of them 
at the moment.  Various environment variables are listed in the manual for 
disabling optimizations and in the configuration help description.

-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

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