[gmx-users] water molecule can not be settled, what could be the cause?

[Jian Zou]

Tsjerk wrote:
> One reason for this may be that a water molecule is
> trapped inside a protein. It can start to oscillate and at
> a certain moment jump to overlap with the protein, causing
> the failure.
> If you find a single water molecule buried
> inside your protein after using genbox, you may want to
> remove it (or several single water molecules).
> It is also possible that you're suffering from problems related to
> the box, for that it's necessary to know the box type you
> use, the version of gromacs and the .mdbp file you use.

> But first try the other thing :)
>
> Hope it helps,
>
> Tsjerk
>
Thank you very much, Tsjerk.
I use gmx-3.2.1, water box is rectangular.
I will check the structure.


I checked the mailing list and found some people have encounter similar 
problems before.

David (if I'm right) ever said that this problem may be caused by that some 
water molecule is surrounded by hydrophobic surface, and starts to spin with 
a high velocity.

Anton (if I'm right) ever made a comment that this high velocity is a 
physical phenomenon and is  not well described by the MD because of the lack 
of energy transfer between the fast-moving water molecule and its 
surrounding environment.
Is the problem as Anton described still unsolved by MD algorithms?


Thank you so much for your comment in advance.


Regards,

Jian Zou