On Wed, Apr 1, 2015 at 10:42 AM, Kshatresh Dutta Dubey <kshatresh.gmail.com>
wrote:
> Hi Jason,
>
> Thanks for your reply.
>
>
> > It's unclear what you mean by "dynamics". There is no single definition
> > of what "dynamics" is -- do you mean that this rotation causes particular
> > behavior? Like the ligand adjusting its orientation in the binding
> > pocket? A different dihedral angle rotating? Does it eject the ligand
> > from the binding pocket?
> >
>
> Here ligand dynamics means 'transnational motion', it is going out of the
> pocket.
>
> >
> >
>
> > You need to form a concrete definition of what you mean by "dynamics" --
> > one that you can describe using various geometric measurements (mean
> > squared displacement, a different torsion angle, COM distance from some
> > residue or point in the binding pocket, atomic fluctuations, etc.). Then
> > it could be as simple as plotting the time series of the Chi2 dihedral
> > angle alongside the time series of whatever measurement you made to
> > quantify "ligand dynamics".
> >
>
> I have calculated the chi2 dihedral for phenyl ring and variations in COM
> for ligand with respect to an almost stationary point in the complex.
>
>
> > You can also compute a cross-correlation function between the two time
> > series if you need a more quantitative measure of the correlated motion
> > (rather than just demonstrating that a particular event in one curve
> > follows immediately from an event in the other one).
> >
>
> Yes, I want to calculate a cross-correlation between these two data sets
> (chi2 dihedral and distance change of ligand). I went through the manual
> and it seems that 'corr' command will be suitable for it. However, I am not
> sure about the dataset1 and dataset2. Should I use cpptraj out files for
> dihedral and COM directly as dataset1 and dataset2, directly?
>
You should be able to do this all within the same cpptraj run. The
dihedral command creates a data set for the dihedral you measured. The
distance command also creates a data set. (You are free to name any of
these data sets.)
Then specify the names of those two data sets to the corr command.
Alternatively, you can create the data sets via the "readdata" command in
cpptraj if you do not want to re-run the calculation, and use the "corr"
command on those data sets. Or, of course, you can write a script to
compute the cross-correlation function between the data sets yourself
(using numpy, for instance).
HTH,
Jason
--
Jason M. Swails
BioMaPS,
Rutgers University
Postdoctoral Researcher
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Apr 01 2015 - 08:00:07 PDT
