[gmx-users] questions - ED analysis

[Lakshmi Padmavathi]

Hi all,

i want to analyze molten globule formation of my protein. so I adopted the
protonated states from pka calculations and did MD simulations(10ns) at pH 7
(native), pH3, pH4 and pH2. I did ED analysis to identify the collective
motions.

1. Can I use the crystal structure as reference structure to do g_covar?
   Fitting to the average structure or fitting to the reference structure.
Which one is better?

2. Can I believe the overlap, cosine content and RMSIP values which I got
from g_covar and g_anaeig by fitting to the reference structure? because I
got low cosine values when I fit to the reference structure (crystal str)
and higher cosine values when I fit to the average structure. 

3. Does it make sense to do combine ED analysis to see changes among
different pH simulation?  If yes, which structure I have to use to fit for
the combined trajectory? (crystal or average)

I hope some of you could give suggestion for me.

Thank you very much

Best regards,
lakshmi