You haven't given us much info. First, you need to ensure that these basins
are all reliable and reproducible. Free energies from straight md are
notoriously unreliable and it's easy to over interpret the results. Make
sure you can physically justify them as well and they are not just noise.
To get back to your question, you ask about the "initial stage of
simulation". What exactly takes place in this simulation? What are you
doing? What do you mean "change while ligand binding"? Are you forcing
binding to occur?
One way to get the insight you're looking for is to map the time dependence
of the projection of each snapshot onto the pca modes. That way you can map
basins to structures. Explaining what's going on would then take some
structural biology insight, knowledge of your specific system, and most
likely design of followup data analysis to confirm that what you see in the
snapshots is actually a well sampled property in the MD run and not just an
artifact of looking at a subset of the snapshots. Then you would use this
information to compare to experimental data, keeping in mind all of the
potential weaknesses of your simulation. Ideally you will make specific
predictions that are then confirmed through experiments.
On Nov 26, 2014 6:07 AM, "anu chandra" <anu80125.gmail.com> wrote:
> Dear Amber users,
>
> I am working with MD simulation of protein -ligand interaction. In order to
> get information about subtle change in protein while ligand binding, I have
> carried out principal component analysis (PCA). I have also drawn the free
> energy landscape using PC1 and PC2, which came up with five low-energy
> basins. is there a way to identify which basin will be encountered at the
> initial stage of simulation?
>
> Many thanks,
>
> Anu
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Nov 26 2014 - 09:30:02 PST
