[gmx-users] RE: RE : FEP : separating components of dgdl (Berk Hess)

[Berk Hess]

----------------------------------------
> Date: Wed, 30 Jan 2008 11:59:30 +0100
> From: michael.bon at cea.fr
> To: gmx-users at gromacs.org
> Subject: RE: RE : FEP : separating components of dgdl (Berk Hess)
> 
> 
> Hi,
> 
> After some tedious work to locate the problem, I just realized that the term for bonds is missing in the log file !
> It could account for the difference, but in fact another error occured :
> 
> (Note : I am using amber99 ports )
> 
> In my system, I defined an atom DUM with just the same parameters as H. When I do the transformation DUM->H, I have dVdl=0, which is OK.
> If I change the parameters for bonds, that's to say in my .top file :
> 
>    [ bondtypes ]
>       NC  DUM   1  0.101  0;
>       NC  H     1  0.101  363000;
> 
> (DUM and H are bound to a NC)
> 
> I still get dVdl = 0 ?! There should be a change due to this difference ( I did a run at lambda = 0.5 ) !
> 
> Moreover, when I cancel one of those two lines by adding a ";" at the beginning, I have an error message when using grompp saying "no default bond types..". So the program does not ignore those lines and needs them.
> 
> And if I change a parameter for angle, for example
> 
>    [ angletypes ]
> 
>      C  NC  DUM  1  120.00  0.
>      C  NC  H    1  120.00  400.
> 
> I do have a difference in dVdl
> 
> 
> What is happening ?

I guess you are constraining all bonds and therefore you have no bond potentials.

Berk.

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE!
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/