[gmx-users] Re: Velocity autocorrelation function of water using g_velacc

[Andrew DeYoung]

Hi,

I just want to say thank you to everyone who took the time to respond to my
post about the VACF of water.  I really, really appreciate it.  

I made a very silly mistake: I equilibrated the system for 5 ns... only to
throw it all away!  It was a very big mistake on my part, and without your
help, to be honest, I may have blindly gone on making that mistake for a
while, although I hope not.  So, many thanks for your help, and thank you
also for being patient with me.  

Initially my diffusion coefficient via the integral of the VACF (by the
Green-Kubo relation) did not yield the same result as the diffusion
coefficient via the mean square displacement... because I was wrongly
integrating the normalized VACF!  So just in case there are any other
novices reading this mailing list, please don't make the same mistakes I
have.  To obtain the diffusion coefficient from the VACF, be sure to
integrate the UNnormalized VACF, using the option -nonormalize in g_velacc.

Thanks again, 

Andrew DeYoung
Carnegie Mellon University