[gmx-users] Re: Some questions about genbox and EM (Tsjerk Wassenaar)

Hu Zhongqiao zhongqiao_hu at nus.edu.sg
Tue Jan 2 15:05:26 CET 2007

Thanks, Tsjerk
> >1) Do I need to use genbox to add H2O twice?
>> In the past, I first used genbox to add H2O and then used genion to
>> counterion. After that is EM. But now I find, if I use genbox again
>> genion or even after EM, there are some additional, though very few,
>> molecules will be added into system. So I want to ask if the method
to use
>> genbox twice is more reasonable?

>Assuming that you mean placement inside a protein, I don't think it is
>necessary. For the solvent itself or the interface, it really doesn't
>matter. During EM, or even during MD, some cavities may be generated
>which could be filled with a water molecule, but that doesn't mean
>they should be occupied. Of course, it would be interesting to test
>whether your results improve, i.e. one or the other leads to a more
>stable system. In that case, you should also try the placement of
>water by optimization of the orientation (which isn't done using
>genbox and is much more costly).
I am running MD simulation in some protein crystals. I always find that
the water density in some big pore within protein crystals is ca. 950 or
even 920 kg/m^3 at 300K and SPC model, obvioulsy less than 1000 kg/m^3
(bulk density). I think it will be more reasonable if water density is
closer to bulk density, right?
>> 2)One or two EM?
>> Some users mention that one EM will be done after adding H2O and the
>> will be done after adding counterion. In the past, I just did one EM
> >after adding counterion. Is it more reasonable to do one EM right
>> adding H2O?
>It's not likely to be necessary. EM (prior to MD) is just to obtain a
>reasonable structure which can be used to start an MD run with,
>without exploding due to bad contacts/strained geometries.
You are right in this point. I totally agree with you.
> >3) bond-length constraint in EM or not?
>> In the past, I did EM without bond-length constraint. But lately one
>> who is obviously more expericed in MD than me, told me that I should
do 2
> >EM, 1st with bond-length constraint and 2nd without it. Is it more
>> reasonable?
>Again, not necessary, see point 2.
I wonder if the structure of the system under the study is changed
during EM. For example, in my system consisting protein molecules, water
and counterion, some bond lengths in protein molecules will be changed
if I do EM without bond-lenght constraint. In this case, some changed
bond lengths in EM will be kept constant in equilibration and production
MDs because I use bond-length constraint. I doubt its validaty. So I am
thinking if it is more reasonable to carry out EM with bond-length
>>Any comments are welcome.

>I hope this helps.

Zhongqiao Hu
Lab E5-04-27
Tel: 65161946(O)
Department of Chemical and Biomolecular Engineering
National University of Singapore
117576, Singapore
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20070102/6dd5bd46/attachment.html>

More information about the gromacs.org_gmx-users mailing list