[gmx-users] Implicit solvent - Gromacs 4.5.4

[Vedat Durmaz]

thanks justin for clearing up these issues ..


Am 28.02.2012 15:50, schrieb Justin A. Lemkul:
>
>
> Vedat Durmaz wrote:
>>   actually, as far as i know, the -DFLEXIBLE argument has no effect 
>> on simulations without explicit water ?!
>>
>
> The #ifdef FLEXIBLE block does only affect explicit water, unless the 
> topology is somehow modified to use this type of term (probably not).  
> It shouldn't be used for MD anyway, as all water models were 
> parameterized to be rigid.
>
>> i just copied the "define" line from my explicit simulation systems 
>> .. so, the only define value, that has a meaning is -DPOSRESHELIX.
>>
>> i didn't use constraints for the first time in my life since implicit 
>> solvent simulations incorporate a much smaller number of atoms and 
>> are therefore not that expensive computationally. i always use this 
>> setting for my explicit simulations only in order to increase the 
>> step size to 2fs with a clear conscience ...
>
> One should also consider the representation of the physical model.  It 
> can be argued that rigid bonds are a more realistic representation of 
> ground state bonds, in addition to making the simulations more efficient.
>
> -Justin
>