[gmx-users] REMD and T autocorrelation

[Mark Abraham]

----- Original Message -----
From: elena.mol at mail.com
Date: Tuesday, August 3, 2010 0:55
Subject: [gmx-users] REMD and T autocorrelation
To: gmx-users at gromacs.org

 >                             <!-- 		@page { size: 8.5in 11in; margin: 0.79in } 		P { margin-bottom: 0.08in } 	-->        Dear gmx-ers, I have a question regarding replica exchange and temperature autocorrelation..   > I want to calculate the temperature (T) autocorrelation function from a REMD simulation.

OK, so do you want T autocorrelation of trajectories of constant index, or constant ensemble? The former requires a demux.

    > Is it ok to “directly” use g_energy on the .edr files of my trajectory in order to obtain an .xvg file with T vs. t, and then g_analyze on this .xvg files to obtain the autocorrelation function? (as one would do on a “standard” MD without replica exchange)     > Or do I need to “demux” the trajectory files with demux.pl to really monitor each replica in its temperature “jumps”? But, in this latter case..... the procedure described in  http://www.gromacs.org/Documentation/How-tos/REMD   > post-processing acts on .xtc files (through the use of .log files).  The .edr files remain unchanged.  Am I missing something trivial?.......     > What am i monitoring if i do g_energy + g_analyze on "raw" (not demuxed) .edr files from a REMD simulation?......

demux.pl writes an index file that trjcat can use to demux, however eneconv does not have the same functionality available, unfortunately. I'd guess that it shouldn't be hard to implement...

Mark

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20100803/c4f62be9/attachment.html>