[gmx-users] large charge group

[Berk Hess]

No, it does not need to be integer.
With PME it does not matter at all.
With cut-off or reaction-field it is better to have neutral
charge groups when possible, but not at the cost
of making enormous charge groups.

Charge groups should preferable consist of 2, 3 or 4 atoms
which are very close together in space.

Berk.

Date: Thu, 20 Nov 2008 09:29:46 +0800
From: qiaobf at gmail.com
To: gmx-users at gromacs.org
Subject: Re: RE: [gmx-users] large charge group












Hi Berk,
 
Should the total charge for one charge group 
be necessarily an integer, such as 0, 1, -1?
 
If I understand correctly, the answer is yes. 
However in this case, the large charge group always happens 
except opls-aa force field.
 
regards,
 



发件人： Berk 
Hess
发送时间： 
2008-11-20 00:21:07
收件人： Discussion list for 
GROMACS users
抄送： 
主题： RE: [gmx-users] large 
charge group
 
Hi,

The charge groups "only" affect the 
neighbor searching,
which in practice means they affect everything.

In 
Gromacs atoms in two different charge groups will only interactions
when the 
center of geometry of the two charge groups is within rlist.

The worst 
case scenario, which has happend to several people, is:
A molecule with a 
radius of about 0.5 nm is one charge group.
If the cut-off is 1 nm, two 
molecules will only really start interacting
when they already 
overlap.

In 4.0 I put a check on large charge groups.
If you have 
charge groups of more than 10 atoms, grompp will now
warn you with:
"The 
largest charge group contains %d atoms.\n"
"Since atoms only see each other 
when the centers of geometry of the charge groups they belong to are within the 
cut-off distance, too large charge groups can lead to serious cut-off 
artifacts.\n"
"For efficiency and accuracy, charge group should consist of a 
few atoms.\n"
"For all-atom force fields use: CH3, CH2, CH, NH2, NH, OH, CO2, 
CO, etc.\n",

Berk

> Date: Wed, 19 Nov 2008 10:01:01 
-0600
> From: sunnytov at gmail.com
> To: gmx-users at gromacs.org
> 
Subject: [gmx-users] large charge group
> 
> Hello All
> 

> Recently I have figured out the biggest problem that made my system 
unstable.
> I have parameterized a molecule pretty similar to octane. A 
single
> charge group used to be assigned to this molecule. With 1 nm 
cut-off
> distance for LJ interaction and the direct space for 
electrostatic
> interaction, the system underwent severe volume 
fluctuation and at the
> end crashed due to enormously large short range 
coulomb interaction
> energy. The problem was temporarily resolved by 
using greater cut-off
> distance such as 1.4 nm. But this fixation was not 
perfect when I put
> this molecule into more complex system. Regardless of 
the cut-off
> distance even to 1.8 nm did not solve the problem. The 
simulation
> always terminated due to large coulomb energy even though no 
direct
> contact or any weird structure was observed at that 
snapshot.
> 
> With the help of many experts, I have made a single 
charge group into
> three smaller charge group, which magically fixed all 
the problem that
> I have had for almost 2 years. I know that charge group 
is only used
> for neighbor search. Can anybody tell me how the size of 
charge group
> can influence the interaction between atoms depending on 
the
> complexity of the system and cut-off distance??
> 
> 
Thank you in advance
> 
> Sunjoo
> 
_______________________________________________
> gmx-users mailing list 
gmx-users at gromacs.org
> 
http://www.gromacs.org/mailman/listinfo/gmx-users
> Please search the 
archive at http://www.gromacs.org/search before posting!
> Please don't 
post (un)subscribe requests to the list. Use the 
> www interface or send 
it to gmx-users-request at gromacs.org.
> Can't post? Read 
http://www.gromacs.org/mailing_lists/users.php



Express yourself instantly with MSN Messenger! MSN 
Messenger
_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today it's FREE!
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20081120/f1b4224a/attachment.html>