[gmx-users] g_rotacf, order parameter S2 problem

[Berk Hess]

Hi,

I would say rotational acf's can be negative.
In most cases this will indicate bad sampling.
This is not suprising considering the extremely long times you printed.
I guess you get positive values for shorter correlation times.

Berk

> From: X.Periole at rug.nl
> Subject: Re: [gmx-users] g_rotacf, order parameter S2 problem
> To: gmx-users at gromacs.org
> Date: Wed, 14 Jan 2009 12:23:11 +0100
> 
> On Sun, 11 Jan 2009 12:50:00 +0800
>   Dechang Li <li.dc06 at gmail.com> wrote:
> > Dear all, 
> > 
> >    I used g_rotacf to calculate the order parameter (S2, N-H bond in 
> > main chian). Thanks to Xavier Periole's advice, I make the index file
> > and obtained some results. However, some results I got made me puzzling
> > that the order parameter is negative! Here is one of my results:
> > 
> > @    title "Rotational Correlation Function"
> > @    xaxis  label "Time (ps)"
> > @    yaxis  label "C(t)"
> > @TYPE xy
> > ...
> > 119944.000    -0.16902
> > 119946.000    -0.16845
> > 119948.000    -0.16812
> > 119950.000    -0.16845
> > 119952.000    -0.16834
> > 119954.000    -0.16825
> > 119956.000    -0.16875
> > 119958.000    -0.16823
> > 119960.000    -0.16850
> > 119962.000    -0.16885
> > 119964.000    -0.16892
> > 119966.000    -0.16924
> > 119968.000    -0.16885
> > 119970.000    -0.16928
> > 119972.000    -0.16889
> > 119974.000    -0.16901
> > 119976.000    -0.16866
> > 119978.000    -0.16901
> > 119980.000    -0.16895
> > 119982.000    -0.16953
> > 119984.000    -0.16932
> > 119986.000    -0.16961
> > 119988.000    -0.16947
> > 119990.000    -0.16890
> > 119992.000    -0.16952
> > 119994.000    -0.16986
> > 119996.000    -0.16970
> > 119998.000    -0.16941
> > 120000.000    -0.16908
> > 
> > 
> > 	The command I used is : g_rotacf -s Order.tpr -f Order.trr -d -n index.ndx 
> >-P 2 -nice 0 .
> > Actually, when I used the option "-P 1", the negative results appear 
> >similarly. 
> > 	Before I calculated the S2 parameter, I removed the rotation and 
> >translation movement of 
> > the molecule using the tool trjconv. 
> > 	Is that the negative results reasonable? 
> No! the order parameter must be between 0 and 1. Considering the length
> of your simulation (120 ns) most of the NH vectors will have sampled
> enough for you to get a reasonable value. However some of them will not
> and it also depends on the topology of your protein.
> You should look at the curve and try to make sense with its shape.
> Undulations are typical of lack of convergence. Note that in some cases
> it is very difficult to obtain a smooth curve. Eg. we had to run ~1000
> cases and average them to get a acceptable curve. You do not need to
> do that but be aware that it is not always trivial to get a C(t) converged.
> Anyways, a negative value without undulations would indicate that you
> did something wrong somewhere. May be in the fitting procedure.
> 
> XAvier.
> > 
> > 
> > 
> > Best regards,
> > 2009-1-11
> > 
> > 
> > ========================================= 
> > Dechang Li, PhD Candidate
> > Department of Engineering Mechanics
> > Tsinghua University
> > Beijing 100084
> > PR China 
> > 
> > Tel:   +86-10-62773574(O) 
> > Email: lidc02 at mails.tsinghua.edu.cn
> > =========================================
> > 
> >  
> > 
> > _______________________________________________
> > 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
> 
> -----------------------------------------------------
> XAvier Periole - PhD
> 
> - Molecular Dynamics Group -
> Computation and NMR
> University of Groningen
> The Netherlands
> -----------------------------------------------------
> _______________________________________________
> 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

_________________________________________________________________
See all the ways you can stay connected to friends and family
http://www.microsoft.com/windows/windowslive/default.aspx
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20090114/26c0429a/attachment.html>