[gmx-users] settle for other solvents

[Erik Lindahl]

On Monday, Feb 17, 2003, at 23:41 US/Pacific, Christoph Freudenberger 
wrote:

> Hi there,
>
> Is it possible to use the settle procedure for other solvents than spc?
> CH2Cl2 should be possible (with a united CH2-group) because it has the
> symmetry like the water molecule.
> But what about linear molecules like Acetonitrile: Me-C=N?
> Or 4-atom models like DMSO?
>       Me
>      /
>   O=S
>      \
>       Me
>

Hi,

No - Settle is currently restricted to water-like three-particle 
molecules, i.e. non-linear and symmetrical (atoms 2 & 3 must be 
identical).  It is possible to cast it in a slightly more general form 
that doesn't assume the symmetry (still limited to three atoms, 
though), but I suspect you would run into some nasty singularities for 
a linear molecule. (This is slightly more costly, though, so it 
shouldn't be done for water).

Settle relies on analytically solving the constraints, so the 
complexity will increase rapidly with the number of atoms involved. I 
would *guess* that it is about as cheap to use numerical algorithms 
with four or more atoms. In practice it's probably not worth the 
coding; just try to use a normal bond-constraining algorithm like 
shake/lincs!

Cheers,

Erik








> Best regards?
> -- 
> Christoph Freudenberger
> Abt. Organische Chemie I AK Siehl - Uni Ulm -Tel: ++49-731-502-2785
>
> _______________________________________________
> gmx-users mailing list
> gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please don't post (un)subscribe requests to the list. Use the www 
> interface or send it to gmx-users-request at gromacs.org.