[gmx-developers] introduce effective potential

[Nguyen Hoang Phuong]

On Mon, 19 Apr 2004, Berk Hess wrote:

> >Dear All,
> >
> >I would like to replace the dihedral potential energy function (formed
> >by 4 atoms i,j,k,l) by the effective potential
> >
> >U_eff(\theta_{ijkl}) = KT*q/(q-1)*log(1 + KT*(q-1)*[U(\theta_{ijkl}) +
> >epsilon])
> >
>
> It seems to me that the second factor KT in this formula should be (1/KT).

Thanks a lot. Yes, you are correct.

>
> >where q and epsilon are the parameters, U(\theta_{ijkl}) is the original
> >potential. This effective potential is smoother (depends on value of q)
> >than the
> >  originial
> >one and therefore the sampling is enhanced.
> >
> >This is equivalent to introduce the scale factor:
> >
> >q/(1 + KT*(q-1)*[U(\theta_{ijkl}) + epsilon])
> >
> >into the forces which acting on the atoms i,j,k,l
> >
> >To this end, I have modified to code as follow (in bondfree.c)
>
> The modification is not correct, because the shift forces for the virial
> calculation are not modified and thus the pressure will be incorrect.

Could you tell me the subroutine which is needed to be done to get
correct the pressure?

>
> You should scale ddphi just before the do_dih_fup call or F* in dopdihs.
>
> Note that you did not modify the potential value and thus your potential
> energy will be incorrect.
> If you want it to be correct you need to apply your conversion formula
> to vpd (or V* in dopdihs).

OK, that is fine.

At the moment, whenever I try with each value of q and epsilon, I have to
compile the gromacs. It is very time consumming and of course it is a very
bad way. How can I introduce these parameters into the *mdp file?

Thanks again for your help.

Phuong

>
> Berk.
>
> _________________________________________________________________
> Talk with your online friends with MSN Messenger http://messenger.msn.nl/
>
> _______________________________________________
> gmx-developers mailing list
> gmx-developers at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-developers
> Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-developers-request at gromacs.org.
>