<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7651.59">
<TITLE>Re: [gmx-users] dynamic cross correlation map</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<P><FONT SIZE=2>Hi, all:<BR>
<BR>
Does anybody know how to normalize the covariance matrix generated by g_covar with option -xpma. For -ascii, these data is the whole covariance matrix, which means the number in the matrix is 3N*3N (N is the number of atoms and I just choose alpha C). Now, I want to get the atomic covariance matrix for each atom pair the sum of the xx, yy and zz. I know -xpma can implement that, but the results is not normalized, it is hard to compared each other. So, I hope someone can tell me what should I do now?<BR>
<BR>
Thanks in advance.<BR>
<BR>
Qi Yan<BR>
<BR>
I think g_covar -ascii is what you are looking for.<BR>
I have a Maple script that does this kind of graph. If you want you can<BR>
contact me off list.<BR>
<BR>
Regards.<BR>
<BR>
Pedro.<BR>
<BR>
2007/9/18, Dhananjay <[EMAIL PROTECTED]>:<BR>
><BR>
> Well, I have done 30ns MD simulation for a protein. The protein consists<BR>
> of 4 loop regions.<BR>
> Using g_cover programme the covariance matrix have been generated and<BR>
> using g_anaeig programme with option -rmsf , along first 8 eigen vectors the<BR>
> rms fluction of c-alpha atoms were plotted.<BR>
> But this is giving me just the information that particular atom is<BR>
> fluctuating along particular direction.<BR>
> I want to know over a 30 ns simulation whether the motions of two atoms or<BR>
> group of two atoms are correlated or anti correlated. For this I want to<BR>
> form a dynamics cross correlation map in which I could get the precise<BR>
> information of group of atoms. The function used for this kind of analysis<BR>
> is<BR>
><BR>
> C(i,j) = < delta r(i) * delta r(j) > / sqrt < sqr(delta r(i) ) > . sqrt <<BR>
> sqr(delta r(j) ) ><BR>
><BR>
> the positive C(i,j) -> motions are correlated<BR>
> the negative C(i,j) -> motions are anti-correlated<BR>
><BR>
> I want to plot a 2-D map indicated correlated and anti-correlated motions.<BR>
><BR>
><BR>
> My question is whether it is possible in GROMACS to plot this king of map<BR>
> or could you please suggest any other free software which could read the<BR>
> trajectories generated by GROMACS and plot the map.<BR>
><BR>
> Again thanking you in advance ......<BR>
><BR>
><BR>
> -- Dhananjay<BR>
><BR>
><BR>
><BR>
><BR>
> On 9/18/07, David van der Spoel <[EMAIL PROTECTED]> wrote:<BR>
> ><BR>
> > Dhananjay wrote:for each atom pair the sum of the xx, yy and zz covariances<BR>
> > > This may be bit different question from main theme.<BR>
> > ><BR>
> > > I want to form dynamic cross correlation map as I have trajectories<BR>
> > > generated by GROMACS 3.3.<BR>
> > ><BR>
> > > Please suggest me free software which can read the .trr file for<BR>
> > > plotting DCCM (dynamic cross correlation map )<BR>
> > try g_covar or otherwise explain in more detail (equations) what you<BR>
> > want to do.<BR>
> ><BR>
> > ><BR>
> > > Thanking you in advance....<BR>
> > ><BR>
> > ><BR>
> > ><BR>
> > > -- Dhananjay<BR>
</FONT>
</P>
</BODY>
</HTML>