[gmx-users] Free energy calculations - desolvation energy of Na+

[Justin A. Lemkul]

eva.pluharova at marge.uochb.cas.cz wrote:
> Hi Justin,
> 
> thank you for suggestion about changing the integrator, but I would like
> to know why it is not possible to sample properly by md under these
> circumstances.
> 
> 1. Do you know any papers about sampling efficiency of systems with almost
> or completely switched off particles? How does it depend on the kind of
> thermostat or barostat?
> 

I don't know about papers, but there are several relevant discussions in the 
list archive, i.e.:

http://lists.gromacs.org/pipermail/gmx-users/2003-June/006054.html

> 2. Note, that you can switch off the LJ particle also by creating B
> topology with sigma and epsilon equal to 0 in the .top file, not using
> couple-moltype. Doing so, there is no warning, but these approaches are
> equivalent. This seems like internal inconsistency in Gromacs.
> 

Sorry, can't comment here.  Topology manipulation is outdated as of version 4.0, 
but I guess it still works.  Maybe some checks need to be implemented in this 
case.  It certainly seems like there can be artifacts if other .mdp parameters 
are not set properly.

-Justin

> Thanks again for any hints.
> 
> Best,
> 
> Eva
> 
> 
>>>  Hello all,
>>>
>>>  I am trying calculate desolvation free energy of Na+ in water using
>>> option
>>>  couple-moltype, not by creating B-topology in the .top file.
>>>
>>>  At first, I switched off coulombic interaction using:
>>>
>>>  couple-moltype           = Na+
>>>  couple-lambda0           = vdw-q
>>>  couple-lambda1           = vdw
>>>
>>>  Then I tried to switch off LJ interaction using couple-lambda0 = vdw
>>> and
>>>  couple-lambda1 = none, but obtained this warning:
>>>
>>>  WARNING 1 [file equil_NVT000.mdp, line 95]:
>>>    For proper sampling of the (nearly) decoupled state, stochastic
>>> dynamics
>>>  should be used
>>>
>>>  This is strange, because my .mdp file has only 94 lines.
>>>
>> Regardless of the line number, the error message is pretty explicit - you
>> probably have "integrator = md" in your .mdp file when you should have
>> "integrator = sd."
>>
>> -Justin
>>
>>>  Thanks in advance for any help or suggestion.
>>>
>>>  Best,
>>>
>>>  Eva
>>>
>>>  For convenience, I am pasting weak coupling algorithm section and free
>>>  energy section from my .mdp file. The simulation was performed at
>>> constant
>>>  volume using strong temperature coupling, since it was an equilibration
>>>  run.
>>>
>>>  ; OPTIONS FOR WEAK COUPLING ALGORITHMS
>>>  Tcoupl                   = v-rescale
>>>  tc-grps                  = system
>>>  tau_t                    = 0.1
>>>  ref_t                    = 300
>>>  nsttcouple               = 1
>>>
>>>  ; Free energy control stuff
>>>  free_energy              = yes
>>>  init_lambda              = 0.0
>>>  delta_lambda             = 0
>>>  foreign_lambda           = 0.1
>>>  sc-alpha                 = 0.7
>>>  sc-power                 = 1
>>>  sc-sigma                 = 0.3
>>>  nstdhdl                  = 10
>>>  separate-dhdl-file       = yes
>>>  dhdl-derivatives         = yes
>>>  dh_hist_size             = 0
>>>  dh_hist_spacing          = 0.1
>>>  couple-moltype           = Na+
>>>  couple-lambda0           = vdw
>>>  couple-lambda1           = none
>>>  couple-intramol          = no
>>>
>> --
>> ========================================
>>
>> 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
>>
>> ========================================
> 

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

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

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