<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 10pt;
font-family:Tahoma
}
--></style>
</head>
<body class='hmmessage'>
Hi,<br><br>Unfortunately constraints can not be applied to virtual sites and grompp apparently<br>does not not check for this. I will add a check.<br>Constraints between virtual sites can lead to very complex constraint equations<br>between the masses involved. Thus the general case if difficult to implement.<br>Your cases seems quite simple and could be done with the pull code,<br>if it would support more than one reference group, which I was thinking to implement.<br><br>How many molecules do you have?<br><br>Berk<br><br>> Date: Mon, 6 Dec 2010 11:43:45 +0100<br>> From: fritsch@mpip-mainz.mpg.de<br>> To: gmx-users@gromacs.org<br>> Subject: [gmx-users] Constraining two virtual sites<br>> <br>> Hi everybody,<br>> <br>> <br>> I have some trouble setting up topology for my, I admit, quite unusual <br>> system. I need to constrain the center of masses of two groups of atoms <br>> to a fixed 'bond' length. Since I have to do this for every molecule I <br>> cannot use the 'pull constraint' of the free energy code.<br>> <br>> Thus in my topology I defined a com vsite for each group and tried to <br>> apply a constraint between them. The atoms within each group are <br>> connected by harmonic springs forming a ring polymer (for those of you <br>> who know it: this is supposed to become the path integral representation <br>> in the end...)<br>> The topol.top for a single molecule test system looks like this:<br>> <br>> [ defaults ]<br>> ; nbfunc comb-rule gen-pairs fudgeLJ fudgeQQ<br>> 1 1 no 1.0 1.0<br>> [ atomtypes ]<br>> ;name mass charge ptype C6 C12 ; <br>> sigma epsilon<br>> A 1.00800 0 A 0 0<br>> B 1.00800 0 A 0 0<br>> C 1.00800 0 A 0 0<br>> VS 0.000 0 V 0 0<br>> [ moleculetype ]<br>> ; molname nrexcl<br>> SOL 0<br>> [ atoms ]<br>> ; id at resnr resnm atnm cgnr charge<br>> 1 VS 1 A V1 1 0<br>> 2 VS 1 B V2 2 0<br>> 3 A 1 A A 3 0 <br>> 4 B 1 A B 4 0 <br>> 5 C 1 A C 5 0 <br>> 6 A 1 B A 6 0 <br>> 7 B 1 B B 7 0 <br>> 8 C 1 B C 8 0 <br>> [ virtual_sitesN ]<br>> ;<br>> 1 2 3 4 5<br>> 2 2 6 7 8<br>> [ bonds ]<br>> ; ai aj<br>> 3 4 1 0 1687<br>> 4 5 1 0 1687<br>> 5 3 1 0 1687<br>> 6 7 1 0 1687<br>> 7 8 1 0 1687<br>> 8 6 1 0 1687<br>> [ constraints ]<br>> ; ai aj funct length_A<br>> 1 2 1 0.1633<br>> [ system ]<br>> TEST<br>> [ molecules ]<br>> SOL 1<br>> <br>> The conf.gro is setup such that the constraint is satisfied initially.<br>> <br>> When I run the simulation with LINCS the constraint seem to be not <br>> applied at all, the distance beetween the two vsites increases just as <br>> in free diffusion. With SHAKE the system explodes immediately (without <br>> error message, but every value is inf or nan).<br>> I am running the system with the sd integrator in (vers. 4.5.3) but also <br>> tried md and no thermostat, with no change in results.<br>> <br>> I appreciate any advice on this!<br>> <br>> Thanks,<br>> Sebastian<br>> <br>> <br>> <br>> <br>> <br>> -- <br>> gmx-users mailing list gmx-users@gromacs.org<br>> http://lists.gromacs.org/mailman/listinfo/gmx-users<br>> Please search the archive at http://www.gromacs.org/Support/Mailing_Lists/Search before posting!<br>> Please don't post (un)subscribe requests to the list. Use the <br>> www interface or send it to gmx-users-request@gromacs.org.<br>> Can't post? Read http://www.gromacs.org/Support/Mailing_Lists<br>                                            </body>
</html>