<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html><head><title>Re[2]: [gmx-users] using SETTLE for constraints</title>
<META http-equiv=Content-Type content="text/html; charset=iso-8859-1">
<meta http-equiv="Content-Style-Type" content="text/css">
<style type="text/css"><!--
body {
margin: 5px 5px 5px 5px;
background-color: #ffffff;
}
/* ========== Text Styles ========== */
hr { color: #000000}
body, table /* Normal text */
{
font-size: 9pt;
font-family: 'Courier New';
font-style: normal;
font-weight: normal;
color: #000000;
text-decoration: none;
}
span.rvts1 /* Heading */
{
font-size: 10pt;
font-family: 'Arial';
font-weight: bold;
color: #0000ff;
}
span.rvts2 /* Subheading */
{
font-size: 10pt;
font-family: 'Arial';
font-weight: bold;
color: #000080;
}
span.rvts3 /* Keywords */
{
font-size: 10pt;
font-family: 'Arial';
font-style: italic;
color: #800000;
}
a.rvts4, span.rvts4 /* Jump 1 */
{
font-size: 10pt;
font-family: 'Arial';
color: #008000;
text-decoration: underline;
}
a.rvts5, span.rvts5 /* Jump 2 */
{
font-size: 10pt;
font-family: 'Arial';
color: #008000;
text-decoration: underline;
}
span.rvts6
{
font-weight: bold;
color: #800000;
}
a.rvts7, span.rvts7
{
color: #0000ff;
text-decoration: underline;
}
span.rvts8
{
font-weight: bold;
color: #800000;
}
span.rvts9
{
color: #800000;
}
span.rvts10
{
}
/* ========== Para Styles ========== */
p,ul,ol /* Paragraph Style */
{
text-align: left;
text-indent: 0px;
padding: 0px 0px 0px 0px;
margin: 0px 0px 0px 0px;
}
.rvps1 /* Centered */
{
text-align: center;
}
--></style>
</head>
<body>
<p><span class=rvts6>> Berk wrote:</span></p>
<p><br></p>
<p><span class=rvts6>> Things are a bit confusing here.</span></p>
<p><br></p>
<p><span class=rvts6>> SETTLE versus other constraint algorithm is chosen at the topology level</span></p>
<p><span class=rvts6>> by putting [ settle ] or [ constraints ] sections in your topology.</span></p>
<p><span class=rvts6>> With of the other constraint algorithms (LINCS or SHAKE) is used</span></p>
<p><span class=rvts6>> for the constraints in the [ constraints ] sections is set in the mdp file.</span></p>
<p><br></p>
<p>:yes: That is just the thing I need. Thank you!</p>
<p><br></p>
<p><span class=rvts6>> For future Gromacs versions I would like to have the [ settle ] sections</span></p>
<p><span class=rvts6>> in the water topology replaced with [ constraints ] and let mdrun</span></p>
<p><span class=rvts6>> decide if to use settle.</span></p>
<p><span class=rvts6>> From a user point of view, this will not change much though,</span></p>
<p><span class=rvts6>> since settle is only useful for water and a user (nearly) never</span></p>
<p><span class=rvts6>> writes a topology for water.</span></p>
<p><br></p>
<p><span class=rvts6>> Berk</span></p>
<p><br></p>
<p><span class=rvts6>> Date: Mon, 7 Sep 2009 18:28:05 +1000</span></p>
<p><span class=rvts6>> From: </span><a class=rvts7 href="mailto:Mark.Abraham@anu.edu.au">Mark.Abraham@anu.edu.au</a></p>
<p><span class=rvts6>> To: </span><a class=rvts7 href="mailto:gmx-users@gromacs.org">gmx-users@gromacs.org</a></p>
<p><span class=rvts6>> Subject: Re: [gmx-users] using SETTLE for constraints</span></p>
<p><br></p>
<p><span class=rvts6>> Dmitri Dubov wrote:</span></p>
<p><span class=rvts6>> > Dear GMX'ers,</span></p>
<p><span class=rvts6>> > </span></p>
<p><span class=rvts6>> > I am confused with some basic questions about using SETTLE constraining algorithm in gromacs. </span></p>
<p><span class=rvts6>> > </span></p>
<p><span class=rvts6>> > 1) This solution of SHAKE is used, particularly, for spc water. So spc.itp file contains the [ settles ] section, and there is no another [ constraints ] within it. But the parameter "constraint-algorithm" in /water/grompp.mdp (see the gromacs native tutorial for spc water simulation) is set to "Lincs". What is the REAL treatment of water constraints in this case?</span></p>
<p><br></p>
<p><span class=rvts6>> GROMACS transparently uses the best constraint algorithm that you allow </span></p>
<p><span class=rvts6>> it use on a given molecule. The only way you can *not* use constraints </span></p>
<p><span class=rvts6>> on water with the standard water .itp files is to #define FLEXIBLE, as </span></p>
<p><span class=rvts6>> you will see in those .itp files. Since settle is specified there it is </span></p>
<p><span class=rvts6>> used.</span></p>
<p><br></p>
<p>As I understand, the defining <span class=rvts10>[ constraints ] instead of [ settles ] in topology file gives the way to use both SHAKE and LINCS for water. If you need, of course :-)</span></p>
<p>The last (not least) question. Whether the SETTLE is always the best algorithm for water? Or: Supposed I would get different results using SETTLE and LINCS. Does it mean that the reason is in bad LINCS adjustment, and the run with LINCS should be rejected or refined?</p>
<p><span class=rvts6> </span></p>
<p><span class=rvts6>> The choice of LINCS or SHAKE pertains only to molecules for which there </span></p>
<p><span class=rvts6>> is no other constraint option. You can verify this by doing a quick </span></p>
<p><span class=rvts6>> simulation of pure water specifying either of the above and seeing that </span></p>
<p><span class=rvts6>> the results are identical. IIRC the .log file makes a note of what </span></p>
<p><span class=rvts6>> constraints are being used.</span></p>
<p><br></p>
<p><span class=rvts6>> > 2) From manual the run parameter "constraint-algorithm" has only two (LINCS or SHAKE) alternatives. How can I specify the SETTLE using?</span></p>
<p><br></p>
<p><span class=rvts6>> See above. Looking for SETTLE in the manual index would have helped :-)</span></p>
<p><br></p>
<p><span class=rvts6>> Mark</span></p>
<p><span class=rvts6>> _______________________________________________</span></p>
<p><span class=rvts6>> gmx-users mailing list </span><a class=rvts7 href="mailto:gmx-users@gromacs.org">gmx-users@gromacs.org</a></p>
<p><a class=rvts7 href="http://lists.gromacs.org/mailman/listinfo/gmx-users">> http://lists.gromacs.org/mailman/listinfo/gmx-users</a></p>
<p><span class=rvts6>> Please search the archive at </span><a class=rvts7 href="http://www.gromacs.org/search">http://www.gromacs.org/search</a><span class=rvts8> before posting!</span></p>
<p><span class=rvts6>> Please don't post (un)subscribe requests to the list. Use the </span></p>
<p><span class=rvts6>> www interface or send it to </span><a class=rvts7 href="mailto:gmx-users-request@gromacs.org">gmx-users-request@gromacs.org</a><span class=rvts8>.</span></p>
<p><span class=rvts6>> Can't post? Read </span><a class=rvts7 href="http://www.gromacs.org/mailing_lists/users.php">http://www.gromacs.org/mailing_lists/users.php</a></p>
<p><br></p>
<p>Dmitri</p>
</body></html>