[gmx-users] Re: Chloroform (CHCl3) solvent box for G53a5 force field

[Pablo Englebienne]

>> I'm trying to simulate a small molecule in a chloroform box using the
>> GROMOS G53a5 forcefield. I realized that the parameters for the  
>> solvent
>> are present in the ffG53a5.rtp file, however I could not find a CHCl3
>> solvent box included in GROMACS. I did find, however, a CHCl3 solvent
>> box equilibrated by PeiQuan Chen
>> (http://lists.gromacs.org/pipermail/gmx-users/2003-May/005572.html)  
>> at
>> http://www.gromacs.org/index.php?title=Download_%26_Installation/User_contributions/Molecule_topologies
>> .
>>
>> I also saw mention of a box for the Amber forcefield, that is now
>> included in AmberTools (amber10/dat/solvents/cform/cform.pdb and
>> chcl3_equil.pdb.1). I took this one and I'm equilibrating it to use  
>> it
>> later.
>>
>> In the mean time, I wanted to know if I overlooked something, and  
>> there
>> is a CHCl3 box to use with the GROMOS forcefield?
>>
>
> Not one that is officially distributed.  If it was, it would be the
> /share/gromacs/top subdirectory with other solvent topologies and  
> structures.
> Probably your best bet is to use the one in the User Contributions  
> section,
> unless you feel the need to create your own and start from scratch.

Thanks for confirming this, Justin.

I decided not to use the user-contributed CHCl3 box because the  
topology is not consistent with the GROMOS atom types: the CH is  
united atom (although the mass is 12.01100, it should probably be  
13.01900?), while in G53a5 there are parameters for C, H and Cl. Has  
someone ever used this box successfully?

I was able to equilibrate the box in NVT (100-200ps gives a stable  
simulation), although it exploded at constant pressure. After looking  
at some references on the mailing lists (both GMX and AMBER), I tried  
increasing tau_p from 2.0 to 5.0 and that yielded a stable 100 ps  
simulation, although the system later (continuing for further ~150 ps)  
started to oscillate wildly in temperature and pressure.

What is the effect of increasing tau_p? Besides making the dynamics  
stable, would changing its value affect the outcome of the simulation  
in any other way?

Thanks again!