[gmx-users] dynamic cross correlation map

[Qi Yan]

Hi, all:

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?

Thanks in advance.

Qi Yan

I think g_covar -ascii is what you are looking for.
I have a Maple script that does this kind of graph. If you want you can
contact me off list.

Regards.

Pedro.

2007/9/18, Dhananjay <[EMAIL PROTECTED]>:
>
> Well, I have done 30ns MD simulation for a protein. The protein
consists
> of 4 loop regions.
> Using g_cover programme the covariance matrix have been generated and
> using g_anaeig programme with option -rmsf , along first 8 eigen
vectors the
> rms fluction of c-alpha atoms were plotted.
> But this is giving me just the information that  particular atom is
> fluctuating along particular direction.
> I want to know over a 30 ns simulation whether the motions of two
atoms or
> group of two atoms are correlated or anti correlated. For this I want
to
> form a dynamics cross correlation map in which I could get the precise
> information of group of atoms. The function  used for this kind of
analysis
> is
>
> C(i,j) = < delta r(i) * delta r(j) > /  sqrt < sqr(delta r(i) ) > .
sqrt <
> sqr(delta r(j) ) >
>
> the positive C(i,j) -> motions are correlated
> the negative C(i,j) -> motions are anti-correlated
>
> I want to plot a 2-D map indicated correlated and anti-correlated
motions.
>
>
> My question is whether it is possible in GROMACS to plot this king of
map
> or could you please suggest any other free software which could read
the
> trajectories generated by GROMACS and plot the map.
>
> Again thanking you in advance ......
>
>
> -- Dhananjay
>
>
>
>
> On 9/18/07, David van der Spoel <[EMAIL PROTECTED]> wrote:
> >
> > Dhananjay wrote:for each atom pair the sum of the xx, yy and zz
covariances
> > > This may be bit different question from main theme.
> > >
> > > I want to form dynamic cross correlation map as I have
trajectories
> > > generated by GROMACS 3.3.
> > >
> > > Please suggest me free software which can read the .trr file for
> > > plotting DCCM (dynamic cross correlation map )
> > try g_covar or otherwise explain in more detail (equations) what you
> > want to do.
> >
> > >
> > > Thanking you in advance....
> > >
> > >
> > >
> > > --  Dhananjay
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20080115/866dd81f/attachment.html>