[gmx-users] Variation of angles of simulation cell - regd

[Mark Abraham]

On 9/09/2012 11:41 PM, ramesh cheerla wrote:
> Dear Gromacs users,
>
>                      Recently I have  converted each frame of the
> trajectory to a pdb file using gromacs tool  "g_trjconv"  and then
> have extracted angles from the CRYST1 record of pdb file, to see  how
> they are varying in  simulation. By doing so I have observed that my
> box angles alpha and gamma are not  at all varying through out the
> simulation these values are remains at "90" degrees  and there is  a
> change in beta value.
> I have used "triclinic box"  for simulations of my system with angles
> alpha = 90, beta=120 and gamma=90 and have used "NPT" ensemble with
> following options in mdp file "pcoupltype      = anisotropic  and
> ref_p           = 1.0 1.0 1.0 0.0 0.0 0.0 "
> here my questions are:
> 1)  What might be the probable reason for constant values of alpha and
> beta ( why these values are not changing).

See 7.3.15, which of course you read before using anisotropic coupling. 
You're making life hard by not giving your full .mdp file, because 
you're leaving us to guess about your compressibilities. Do summarise, 
but don't exclude. :-) Presumably your choices are consistent with a 
monoclinic box staying that way.

> 2) Am I following correct protocol to extract angles,  or Is there any
> better way to extract  angles from trajectory.

If you want lattice parameters, that's probably the best way to get them.

> 3) Am I using correct options in .mdp file.

Can't tell.

Mark