<html>
<head>
<style>
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 10pt;
font-family:Verdana
}
</style>
</head>
<body class='hmmessage'>
<br><br>> Date: Thu, 12 Feb 2009 16:49:33 +0100<br>> From: cseifert@bph.ruhr-uni-bochum.de<br>> To: gmx-users@gromacs.org<br>> Subject: RE: [gmx-users] Physical problems of comm removal<br>> <br>> Thanks for your ideas!<br>> <br>> I will discuss the possible solutions with my supervisor.<br>> <br>> To your PS:<br>> This is a good hint, I changed many optimization flags to get a good<br>> scaling but never changed the axis. I created the box with editconf<br>> -princ. It automatically made z the longest axis and x the shortest.<br>> Since many users use PME, it might be a good idea to change the behavior<br>> of editconf -princ.<br>><br><br>Ah, I didn't know this.<br>I changed the order of -princ for 4.0.4.<br><br>Berk<br> <br>> Again: Thank you very much for your help!<br>> Christian. <br>> <br>> On Thu, 2009-02-12 at 16:28 +0100, Berk Hess wrote:<br>> > Hi,<br>> > <br>> > Another crude solution is to divide your protein into two groups along<br>> > the long axis<br>> > and use the pull code to keep the two components perpendicular to the<br>> > axis<br>> > of the vector between the COMs at zero.<br>> > <br>> > PS: if you have PME your system will run faster when you make x the<br>> > long axis.<br>> > <br>> > Berk<br>> > <br>> > > Date: Thu, 12 Feb 2009 16:21:29 +0100<br>> > > From: cseifert@bph.ruhr-uni-bochum.de<br>> > > To: gmx-users@gromacs.org<br>> > > Subject: RE: [gmx-users] Physical problems of comm removal<br>> > > <br>> > > Hi Berk,<br>> > > <br>> > > the rhombic dodecahedron with 1771 nm^3 is still twice as big as my<br>> > > stretched box.<br>> > > <br>> > > This System scales quite well till 256 CPUs and I normally use this<br>> > > amount of CPUs (I have to hurry up with my PhD thesis), so particle<br>> > > decomposition is not an alternative.<br>> > > <br>> > > Are there other alternatives?<br>> > > <br>> > > Thanks so far!<br>> > > Christian<br>> > > <br>> > > On Thu, 2009-02-12 at 16:09 +0100, Berk Hess wrote:<br>> > > > Hi,<br>> > > > <br>> > > > The cleanest solution is using a rhombic dodecahedron:<br>> > > > volume 13.58^3/sqrt(2) = 1771 nm^3<br>> > > > (I don't know why people still consider truncated octahedrons).<br>> > > > <br>> > > > You can try using comm-mode=angular with as comm group only the<br>> > > > protein.<br>> > > > This should not produce significant artifacts when your protein is<br>> > > > large.<br>> > > > However, this will not run with domain decomposition,<br>> > > > only with particle decomposition (mdrun -pd), which scales worse,<br>> > > > but probably not so bad with 4-8 cores.<br>> > > > <br>> > > > Berk<br>> > > > <br>> > > > > Date: Thu, 12 Feb 2009 16:00:30 +0100<br>> > > > > From: cseifert@bph.ruhr-uni-bochum.de<br>> > > > > To: gmx-users@gromacs.org<br>> > > > > Subject: [gmx-users] Physical problems of comm removal<br>> > > > > <br>> > > > > Hi users!<br>> > > > > <br>> > > > > I simulate a stretched protein in TIP4P water in a cubic box<br>> > > > > (6.83x9.60x13.58=890) with pbc. During a 50 ns simulation, my<br>> > > > protein<br>> > > > > must not rotate (otherwise, it will interact with itself). I<br>> > found<br>> > > > three<br>> > > > > possibilities to avoid a self interaction: (1) a bigger box, (2)<br>> > > > > restraints or (3) comm removal.<br>> > > > > <br>> > > > > In details:<br>> > > > > (1) bigger box:<br>> > > > > I need at least a diameter of 13.58 nm in every direction to<br>> > avoid a<br>> > > > > self interaction (because of a possible rotation of the<br>> > protein).<br>> > > > For a<br>> > > > > cubic box, this would be (13.58^3 =) 2504 nm^3. For a truncated<br>> > > > > octahedron still 1928 nm^3, which would still be twice the box<br>> > size<br>> > > > I<br>> > > > > use now. This is not suitable for me.<br>> > > > > <br>> > > > > (2) restraints:<br>> > > > > These are "long-time" runs, where I want to observe domain<br>> > movement,<br>> > > > so<br>> > > > > restraints are not an alternative for my simulation.<br>> > > > > <br>> > > > > (3) comm removal:<br>> > > > > To use my stretched box, I have to use COMM-MODE = Angular. This<br>> > > > mode<br>> > > > > provokes artifacts, if I use it only on my protein and is<br>> > therefore<br>> > > > > forbidden since GMX4.0.3. But what happens, if I use it for my<br>> > whole<br>> > > > > system? My protein has a mass of 142800 a.m.u. and the solvent<br>> > has a<br>> > > > > mass of 998100 a.m.u. Let us assume, that the solvent movement<br>> > > > reduces<br>> > > > > itself, because its movement is undirected and the mass of the<br>> > > > protein<br>> > > > > is high enough to fall into account (which might not be true<br>> > > > because:<br>> > > > > <br>> > > > > v_{comm_removal}= 1 / (m_{solvent} + m_{protein}) \cdot<br>> > > > > \sum_{i(solvent)}^n{(solvent)} ( m_i \cdot v_i) \cdot<br>> > > > > \sum_{j(protein)}^o{(protein)} ( m_j \cdot v_j)<br>> > > > > <br>> > > > > with \sum v_i=0 (because the movement of the solvent is<br>> > undirected)<br>> > > > the<br>> > > > > equation would be:<br>> > > > > <br>> > > > > v_{comm_removal}= 1 / (m_{solvent} + m_{protein}) \cdot <br>> > > > > \sum_{j(protein)}^o{(protein)} ( m_j \cdot v_j)<br>> > > > > <br>> > > > > which weights the movement of the protein with its mass, put<br>> > divides<br>> > > > it<br>> > > > > by the mass of protein+mass, what may result in a to small<br>> > removal<br>> > > > of<br>> > > > > the comm.)<br>> > > > > <br>> > > > > Beside this problem, do I have to expect that there might be<br>> > other<br>> > > > > problems with COMM-MODE = Angular, COMM-GRPS = System and a<br>> > cubic<br>> > > > box<br>> > > > > with pbc?<br>> > > > > <br>> > > > > <br>> > > > > <br>> > > > > -- <br>> > > > > M. Sc. Christian Seifert<br>> > > > > Department of Biophysics<br>> > > > > University of Bochum<br>> > > > > ND 04/67<br>> > > > > 44780 Bochum<br>> > > > > Germany<br>> > > > > Tel: +49 (0)234 32 28363<br>> > > > > Fax: +49 (0)234 32 14626<br>> > > > > E-Mail: cseifert@bph.rub.de<br>> > > > > Web: http://www.bph.rub.de<br>> > > > > <br>> > > > > _______________________________________________<br>> > > > > gmx-users mailing list gmx-users@gromacs.org<br>> > > > > http://www.gromacs.org/mailman/listinfo/gmx-users<br>> > > > > Please search the archive at http://www.gromacs.org/search<br>> > before<br>> > > > 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/mailing_lists/users.php<br>> > > > <br>> > > > <br>> > > ><br>> > ______________________________________________________________________<br>> > > > Express yourself instantly with MSN Messenger! MSN Messenger<br>> > > > _______________________________________________<br>> > > > gmx-users mailing list gmx-users@gromacs.org<br>> > > > http://www.gromacs.org/mailman/listinfo/gmx-users<br>> > > > Please search the archive at http://www.gromacs.org/search before<br>> > 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/mailing_lists/users.php<br>> > > -- <br>> > > M. Sc. Christian Seifert<br>> > > Department of Biophysics<br>> > > University of Bochum<br>> > > ND 04/67<br>> > > 44780 Bochum<br>> > > Germany<br>> > > Tel: +49 (0)234 32 28363<br>> > > Fax: +49 (0)234 32 14626<br>> > > E-Mail: cseifert@bph.rub.de<br>> > > Web: http://www.bph.rub.de<br>> > > <br>> > > _______________________________________________<br>> > > gmx-users mailing list gmx-users@gromacs.org<br>> > > http://www.gromacs.org/mailman/listinfo/gmx-users<br>> > > Please search the archive at http://www.gromacs.org/search before<br>> > 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/mailing_lists/users.php<br>> > <br>> > <br>> > ______________________________________________________________________<br>> > Express yourself instantly with MSN Messenger! MSN Messenger<br>> > _______________________________________________<br>> > gmx-users mailing list gmx-users@gromacs.org<br>> > http://www.gromacs.org/mailman/listinfo/gmx-users<br>> > Please search the archive at http://www.gromacs.org/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/mailing_lists/users.php<br>> -- <br>> M. Sc. Christian Seifert<br>> Department of Biophysics<br>> University of Bochum<br>> ND 04/67<br>> 44780 Bochum<br>> Germany<br>> Tel: +49 (0)234 32 28363<br>> Fax: +49 (0)234 32 14626<br>> E-Mail: cseifert@bph.rub.de<br>> Web: http://www.bph.rub.de<br>> <br>> _______________________________________________<br>> gmx-users mailing list gmx-users@gromacs.org<br>> http://www.gromacs.org/mailman/listinfo/gmx-users<br>> Please search the archive at http://www.gromacs.org/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/mailing_lists/users.php<br><br /><hr />Express yourself instantly with MSN Messenger! <a href='http://clk.atdmt.com/AVE/go/onm00200471ave/direct/01/' target='_new'>MSN Messenger</a></body>
</html>