That appears to work. Thanks!<br><br>-Paul<br><br>Message: 4<br>
Date: Tue, 4 Mar 2008 10:56:16 +0100 (CET)<br>
From: <a href="mailto:hessb@mpip-mainz.mpg.de">hessb@mpip-mainz.mpg.de</a><br>
Subject: Re: [<span class="nfakPe">gmx</span>-<span class="nfakPe">developers</span>] User defined potential between specific<br>
atoms<br>
To: "Discussion list for GROMACS development"<br>
<<a href="mailto:gmx-developers@gromacs.org"><span class="nfakPe">gmx</span>-<span class="nfakPe">developers</span>@gromacs.org</a>><br>
Message-ID:<br>
<<a href="mailto:65059.91.32.170.145.1204624576.squirrel@squirrel.mpip-mainz.mpg.de">65059.91.32.170.145.1204624576.squirrel@squirrel.mpip-mainz.mpg.de</a>><br>
Content-Type: text/plain;charset=iso-8859-1<br>
<br>
>>> I am using the CVS version (downloaded yesterday).<br>
>>><br>
>>> I am trying to include interactions between specific atoms (i.e. atom 1<br>
>>> and<br>
>>> 40 are interacting and atom 30 and 80 are interacting). I am using<br>
>>> "pairs"<br>
>>> to do this. If I look at large systems, where two bodies can<br>
>>> dissociated,<br>
>>> I run into a problem with dynamics load balancing. Since "pairs" is<br>
>>> considered a bonded interaction I need to set "-rdd" to a large value.<br>
>>> If<br>
>>> i do this, then I can not run on many nodes because the domains are<br>
>>> forced<br>
>>> to be too large. Is there an alternate way to include these<br>
>>> interactions<br>
>>> that won't run into this issue with DD? I have looked at the user<br>
>>> defined<br>
>>> potentials, but it appears you can only set the form of the potential,<br>
>>> which<br>
>>> is then used in "pairs".<br>
>><br>
>> Possibly a bond potential type 6 might work, but I can't think of a<br>
>> reason<br>
>> this wouldn't produce the same kind of problem.<br>
><br>
> It would.<br>
><br>
> There are two ways of solving this problem.<br>
> You could make these atoms separate energy groups<br>
> and use the energygrp_table option to supply a tabulated<br>
> non-bonded interaction.<br>
><br>
> You could also simply remove the missing interactions check in the code.<br>
> Since somebody else also asked about this, we might consider<br>
> adding an option to mdrun to turn off this check, or to turn it off<br>
> for interactions other than normal bond, angle and dihedral potentials.<br>
><br>
> Note that your problem is not specific for dynamic load balancing.<br>
> Setting -rdd much larger than your non-bonded cutoff distance<br>
> will kill your performance.<br>
><br>
> Berk.<br>
<br>
I have added an option -noddbc to mdrun which turns off the check<br>
for pair interactions and tabulated bonds that do not generate<br>
exclusions.<br>
Please check if this options does what it should do.<br>
I did not have a proper test system at hand.<br>
<br>
Berk.<br>
<br>
<br>
<br>
<br>
------------------------------<br>
<br>
______________________________<div id="1ens" class="ArwC7c ckChnd">_________________<br>
<span class="nfakPe">gmx</span>-<span class="nfakPe">developers</span> mailing list<br>
<a href="mailto:gmx-developers@gromacs.org"><span class="nfakPe">gmx</span>-<span class="nfakPe">developers</span>@gromacs.org</a><br>
<a href="http://www.gromacs.org/mailman/listinfo/gmx-developers" target="_blank">http://www.gromacs.org/mailman/listinfo/<span class="nfakPe">gmx</span>-<span class="nfakPe">developers</span></a><br>
<br>
<br>
End of <span class="nfakPe">gmx</span>-<span class="nfakPe">developers</span> Digest, Vol 47, Issue 3<br>
*********************************************</div>