[gmx-users] concaternating remd trajectories using trjcat & demux

Justin A. Lemkul jalemkul at vt.edu
Thu Feb 11 02:49:56 CET 2010

Segun Jung wrote:
> Dear gromacs users,
> I am trying to collect trajectories corresponding to each temperature by 
> using trjcat with demux.
> There were similar issues posted earlier, but I do not see the solution 
> on the problem I am facing following:

Like what?  It will save time if you can post links to these "similar issues" to 
avoid posting non-solutions that have already been ruled out.

> I have 64 replicas simulated using namd  in dcd trajectory format and 
> saved them in gromacs format (.trr) using vmd.
> I inspected the gromacs format trajectories by eyes in vmd and looked 
> fine. However, when I tried trjcat -f g*.trr -demux replica_index.xvg,
> the output looked weird. So I tested a small set of the trajectories 
> (using only replicas 0 to 8 and first two frames) and
> noticed that the output does not match to the replica_index.xvg file.

If you ran your simulations with NAMD, how did you generate this file?  In 
Gromacs, one would run demux.pl on the md.log file.  I presume NAMD prints 
different output.

> trjcat -f *g*.trr -demux replica_index.xvg
> replica_index.xvg
> 0 0 1 2 3 4 5 7 6
> 2 1 0 2 3 5 4 6 7

If you analyzed replicas 0 to 8 (inclusive) then you should have an 8 somewhere, 

> I am using Ubuntu (32bit) and the gromas version is 4.0.5.
> 0_trajout.xtc should have the 1st frame from replica 0 and 2nd frame 
> from replica 1, but both frames for 0_trajout.xtc are from the replica 0.

This might go along with my comment above.  If there are nine replicas (0 to 8), 
then there may be some mis-translation of the .xvg file.

> It seems the index file does not cooperate properly with the trjcat and 
> -demux. Does anyone have clue about this?

I know I can attest to demultiplexing working as advertised, so I assume intact 
trajectories with a correct index file should work properly.  You have a few 
variables to deal with: .dcd->.trr translation, however you generated the .xvg 
file, and if you've even told us the right number of replicas, among perhaps 

Also, what does gmxcheck tell you about each of the .trr files?  Do they contain 
what you would expect them to?


> Many thanks,
> Segun


Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080


More information about the gromacs.org_gmx-users mailing list