[gmx-developers] A minor bug in demux.pl

[Yong Wang]

Hi,

We are running a normal REMD with 10 replicas and an exchange frequency of
0.2ps, and tried to use demux.pl to get continuous trajectories. But we
found the demuxed trajectories always become discontinuous after 100ns.
After some frustrating debugging, we found the replica index file
(replica_index.xvg) was not corrected generated by demux.pl. For example,
the output looks like this:

99998.8                 9    2    4    0    5    8    3    7    6    1

99999                   9    2    4    0    8    5    7    3    6    1

99999.2                 9    2    4    8    0    5    7    3    6    1

99999.4                 9    2    8    4    5    0    7    3    1    6

99999.6                 9    2    8    5    4    0    7    3    1    6

99999.8                 9    2    8    5    4    0    7    3    6    1

100000                  9    2    8    4    5    0    7    3    6    1

100000                  2    9    8    4    0    5    7    3    6    1

100000                  2    9    8    0    4    5    7    6    3    1

100001                  9    2    0    8    5    4    7    6    3    1

100001                  9    2    0    8    5    4    7    3    6    1

100001                  2    9    8    0    4    5    3    7    1    6

100001                  2    9    8    4    0    5    3    7    1    6

100001                  2    9    4    8    0    5    7    3    1    6

100002                  2    9    4    0    8    5    7    3    1    6

100002                  2    9    0    4    8    5    3    7    6    1

100002                  2    9    0    8    4    5    3    7    6    1

100002                  2    9    8    0    4    5    3    7    6    1

100002                  2    8    9    4    0    5    3    7    6    1

As you can see, the time information in the first column was confused after
100000 (corresponding to 100ns). So we tried to correct the time
information, and the problem seems solved. In our case, we made minor
modifications on the demux.pl script by simplify replacing "%-20g" by
"%-20.1f". And it works for us!

I don't know if this is the best solution but I am reporting this and hope
it is helpful to others. And the issue can only occur when we use an
exchange frequency smaller than 1ps and run the simulation longer than
100ns. Could you identify if this is a minor bug or just limited to our
case? Thanks a lot!

Regards,
Yong

-- 

==========================
Yong Wang
Postdoc in Lindorff-Larsen Lab
Structural Biology and NMR Laboratory  &  Linderstrøm-Lang Centre for
Protein Science
Department of Biology, University of Copenhagen
Ole Maaløes Vej 5, DK-2200 Copenhagen N, Denmark
https://scholar.google.com/citations?hl=en&user=LblGaZMAAAAJ&view_op=list_works&sortby=pubdate
<https://mailtrack.io/trace/link/01603ac90f476741fe794a895a94d4ee843620d6?url=https%3A%2F%2Fscholar.google.com%2Fcitations%3Fhl%3Den%26user%3DLblGaZMAAAAJ%26view_op%3Dlist_works%26sortby%3Dpubdate&userId=2367320&signature=0578092a37df8130>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-developers/attachments/20190707/583491f4/attachment.html>