[gmx-users] pull code

[Kay Gottschalk]

The profiles you get with pulling simulations are of an irreversible  
(out-of-equilibrium) process. Thus, dissipation has a strong  
contribution, and the forces/energies are higher than for the  
equilibrium process. Furthermore, the reaction coordinate is biased  
due to the choice of the vector you are pulling along - the 'best'  
reaction coordinate for your process can be different, so that again  
you may end up calculating higher forces. Therefore, it might well be  
that the backward process happens spontaneously after 10 ns, even if  
you calculate the high barrier for the forward process. Did you check  
whether the forward process also happens spontaneously after a while?
Cheers,
Kay.

On Jun 23, 2005, at 6:35 PM, Attilio wrote:

> Hi all,
>
> I have a questio concerning the feasibility of a steering dynamics  
> on my
> system. I want to push a small molecule along the minor groove of the
> DNA, so:
>
> 1) I removed all the constraints on bonds (no LINCS or SHAKE)
> 2) I selected as a reference group a piece of DNA, and as a pulling
> group the small molecule.
>
> Maybe I miss something of the warnings in the manual, but I didn't
> understand if the profiles I get are reliable or not (I calculate  
> works
> of ~15 Kcal/mol, but the inverse process -i.e. the movement of the
> molecule in the opposite of the pulling direction- happens  
> spontaneously
> after 10ns!), since I take a part of a molecule as reference group,  
> but
> also I removed all the constraints.
>
> Thanks,
> Attilio
>
>
> -- 
> Attilio Vittorio Vargiu
> PhD Student at SISSA/ISAS
> via Beirut 4
> I-34014 Trieste, Italy
> tel +390403787335   fax +390403787528
>
>
>
> _______________________________________________
> 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.
>