[gmx-users] Re: REMD analysis

francesco oteri francesco.oteri at gmail.com
Fri May 24 12:40:46 CEST 2013


Hi to everybody,
Bharat, maybe i didn't follow exactly the wole tale, but is it possible you
are running xmgrace without the -nxy option?
You are probably visualizing the data related the 1st replica several times!

Francesco


2013/5/24 Mark Abraham <mark.j.abraham at gmail.com>

> On Fri, May 24, 2013 at 10:44 AM, bharat gupta <bharat.85.monu at gmail.com
> >wrote:
>
> > Dear Sir,
> >
> > Thank you for your detailed response to my query. I understood the
> concept
> > of ordered arrangement of ensembles in replica_index.xvg. But I have a
> > doubt, you said that " *At time 4, replicas in ensemble 1 and 2 have
> > exchanged. So replica 0 is now in ensemble 2, which is expressed by 0 in
> > the third column*
> > *of the third row of replica_index.xvg.* "
> >
> > This is fine , as the output of replica_index is :-
> > 4           1    2    *0 *   3    4    5    6    7    8    9   10   11
> 13
> >   12
> >
> > But, i didn't understand this " The same condition is expressed by the
> > first column of the third row of replica_temp.xvg, where you will find 2,
> > also expressing that replica 0 is in ensemble 2 at time 4." Here's the
> > output for replica_temp . The first column third row is 2, its ok, but,
> its
> > shows that replica 0 is in ensemble 1 instead of 2.
> >
>
> No, if the rows of both matrices describe time, and there are two different
> matrices for the same exchange set, then the information described by a
> column must differ, like I said last email. You are applying the same
> interpretation to a column from either matrix.
>
>
> > 4           2    0    1    3    4    5    6    7    8    9   10   11   13
> > 12
> >
> >
> > In addition to this, in my last mail I showed the temp graph for all
> > replicas. (https://www.dropbox.com/s/0c8gp584v1hvlbx/replica_temp.png) .
> > Not all replicas visit all the temperatures, but some of them visit all
> the
> > temperatures. Is it sufficient to move with the further analysis , as in
> > some papers they show that some replicas visit all the temp which means
> > that the sufficient sampling has been achieved. In my case this is true
> for
> > some of the replicas and the average acceptance ratio achieved was 0.22
> ??
> >
>
> I've answered this question several times. Each replica merely visiting
> each temperature means nothing for converged sampling. There's lots of
> literature here, including stuff by me ;-) A balance of replicas visiting
> ensembles is necessary but not sufficient for the kind of replica flow that
> would be necessary for generalized convergence. One can shrug one's
> shoulders at some point and say things are probably as good as they'll get
> for reasonable cost, but your reviewer might disagree with you. Convergence
> of sampling at a single temperature can be assessed in a similar way as for
> non-REMD simulations, caveat that the exchange events pretty much stop you
> using metrics based on correlation time. If you want to know how to do
> things properly, you need to do some reading.
>
> Mark
>
>
> >
> >
> >
> > On Fri, May 24, 2013 at 5:07 PM, Mark Abraham <mark.j.abraham at gmail.com
> > >wrote:
> >
> > > At time 0 we have an set of replicas and an (ordered) set of ensembles.
> > We
> > > could label these however we liked, but for (in)convenience we use
> > 0-(n-1)
> > > for both. The rows of the matrices in the *.xvg files change with time.
> > At
> > > time 2, replicas in ensemble 0 and 1 have exchanged, so replica 0 is
> now
> > in
> > > ensemble 1. At time 4, replicas in ensemble 1 and 2 have exchanged. So
> > > replica 0 is now in ensemble 2, which is expressed by 0 in the third
> > column
> > > of the third row of replica_index.xvg. The same condition is expressed
> by
> > > the first column of the third row of replica_temp.xvg, where you will
> > find
> > > 2, also expressing that replica 0 is in ensemble 2 at time 4. The
> columns
> > > of the two matrices allow you to see either the profile of which
> replica
> > > was in this ensemble at which time, or which ensemble this replica was
> in
> > > at which time.
> > >
> > > Mark
> > >
> > >
> > > On Fri, May 24, 2013 at 8:43 AM, bharat gupta <
> bharat.85.monu at gmail.com
> > > >wrote:
> > >
> > > > Dear Sir,
> > > >
> > > > I tried a lot to understand the meaning and relation between the .log
> > > file
> > > > and relica_index file, but I was not able to break the code. I tried
> to
> > > > look into gmx forum for some clue, but didn't find any. So, if
> possible
> > > can
> > > > you explain it ...
> > > >
> > > > Replica exchange at step 1000 time 2
> > > > Repl 0 <-> 1  dE = -1.067e+00
> > > > Repl ex  0 x  1    2    3    4    5    6    7    8    9   10   11
> 12
> > x
> > > 13
> > > > Repl pr   1.0       .01       .68       .21       .05       .09
> > .26
> > > >
> > > >
> > > > Replica exchange at step 2000 time 4
> > > > Repl ex  0    1 x  2    3    4    5    6    7    8    9   10   11
> 12
> > > 13
> > > > Repl pr        .91       .32       .00       .07       .18       .08
> > > >
> > > > output of replica_index.xvg
> > > >
> > > > 0           0    1    2    3    4    5    6    7    8    9   10   11
> > 12
> > > > 13
> > > > 2           1    0    2    3    4    5    6    7    8    9   10   11
> > 13
> > > > 12
> > > > 4           1    2    0    3    4    5    6    7    8    9   10   11
> > 13
> > > > 12
> > > >
> > > > On Thu, May 23, 2013 at 11:20 PM, Mark Abraham <
> > mark.j.abraham at gmail.com
> > > > >wrote:
> > > >
> > > > > Looked fine
> > > > >
> > > > >
> > > > > On Thu, May 23, 2013 at 4:13 PM, bharat gupta <
> > > bharat.85.monu at gmail.com
> > > > > >wrote:
> > > > >
> > > > > > Sir,
> > > > > >
> > > > > > What about the description of replica_temp file that I posted in
> > last
> > > > > mail.
> > > > > > I think that's correct ... If you can comment on that, I can move
> > on
> > > > with
> > > > > > replica_index file...
> > > > > >
> > > > > >
> > > > > > On Thu, May 23, 2013 at 10:58 PM, Mark Abraham <
> > > > mark.j.abraham at gmail.com
> > > > > > >wrote:
> > > > > >
> > > > > > > It's a demux. One might want trajectories to be at constant
> > > > > temperature,
> > > > > > or
> > > > > > > constant replica. The two files define the (mutually inverse)
> > > > mappings
> > > > > > > between those representations. So one file tells you which
> > replica
> > > is
> > > > > at
> > > > > > > each temperature, and the other which temperature holds each
> > > replica.
> > > > > > > Nobody's ever written down anything about which is which, so
> > like I
> > > > > said
> > > > > > a
> > > > > > > week back, look at the first few exchanges, see how those are
> > > > > represented
> > > > > > > in the files, and decide for yourself which file's columns/rows
> > > have
> > > > > > useful
> > > > > > > information you want to look at. And do write that decision
> down!
> > > :-)
> > > > > > >
> > > > > > > Mark
> > > > > > >
> > > > > > >
> > > > > > > On Thu, May 23, 2013 at 2:55 PM, simula_460 <
> > > > bharat.85.monu at gmail.com
> > > > > > > >wrote:
> > > > > > >
> > > > > > > > I checked the md.log and replica_temp.xvg file , what I
> > > understood
> > > > is
> > > > > > > that
> > > > > > > > the 'x' means swapping and replica are written this way.
> > > > > > > > For eg.
> > > > > > > > Replica exchange at step 1000 time 2
> > > > > > > > Repl 0 <-> 1  dE = -1.067e+00
> > > > > > > > Repl ex  0 x  1    2    3    4    5    6    7    8    9   10
> > 11
> > > > > 12
> > > > > > x
> > > > > > > 13
> > > > > > > > Repl pr   1.0       .01       .68       .21       .05
> .09
> > > > > > .26
> > > > > > > >
> > > > > > > > output in replica_temp file will be
> > > > > > > > 1 0 2 3 4 5 6 7 8 9 10 11 13 12
> > > > > > > >
> > > > > > > > It means that replica 1 at higher temp. exchange with the one
> > in
> > > > > lower
> > > > > > > temp
> > > > > > > > 0.
> > > > > > > >
> > > > > > > > Replica exchange at step 2000 time 4
> > > > > > > > Repl ex  0    1 x  2    3    4    5    6    7    8    9   10
> > 11
> > > > > 12
> > > > > > > 13
> > > > > > > > Repl pr        .91       .32       .00       .07       .18
> > > > .08
> > > > > > > >
> > > > > > > > output in replica_temp file will be
> > > > > > > > 1 0 2 3 4 5 6 7 8 9 10 11 13 12
> > > > > > > > 2 0 1 3 4 5 6 7 8 9 10 11 13 12
> > > > > > > >
> > > > > > > > [order is from low to high temp]
> > > > > > > >
> > > > > > > > But I am not able to understand for replica_index file :-
> > > > > > > >
> > > > > > > > for the above two time steps here's the output :-
> > > > > > > > 0           0    1    2    3    4    5    6    7    8    9
> 10
> > > > 11
> > > > > > 12
> > > > > > > > 13
> > > > > > > > 2           1    0    2    3    4    5    6    7    8    9
> 10
> > > > 11
> > > > > > 13
> > > > > > > > 12
> > > > > > > > 4           1    2    0    3    4    5    6    7    8    9
> 10
> > > > 11
> > > > > > 13
> > > > > > > > 12
> > > > > > > >
> > > > > > > > The time step four is different here, I don't know why ??
> > Ideally
> > > > the
> > > > > > > > output should be same in both files, I suppose ??
> > > > > > > >
> > > > > > > > Also, I tried to plot for each column separately , here I
> want
> > to
> > > > > > clarify
> > > > > > > > that whether each column represents the time evolution of
> each
> > > > > replica
> > > > > > > over
> > > > > > > > time. For eg. the second column should represent the temp
> > > evolution
> > > > > for
> > > > > > > > replica No. 0 wrt to time. Presuming that I understood it
> > > > correctly,
> > > > > I
> > > > > > > > plotted the temp. evolution over time of all replicas
> > separately
> > > .
> > > > > > Here's
> > > > > > > > the replica_temp plot for replicas 0 to 13.
> > > > > > > >
> > > > > > > > https://www.dropbox.com/s/0c8gp584v1hvlbx/replica_temp.png
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > --
> > > > > > > > View this message in context:
> > > > > > > >
> > > > >
> > http://gromacs.5086.x6.nabble.com/REMD-analysis-tp5008199p5008481.html
> > > > > > > > Sent from the GROMACS Users Forum mailing list archive at
> > > > Nabble.com.
> > > > > > > > --
> > > > > > > > gmx-users mailing list    gmx-users at gromacs.org
> > > > > > > > http://lists.gromacs.org/mailman/listinfo/gmx-users
> > > > > > > > * Please search the archive at
> > > > > > > > http://www.gromacs.org/Support/Mailing_Lists/Search before
> > > > posting!
> > > > > > > > * Please don't post (un)subscribe requests to the list. Use
> the
> > > > > > > > www interface or send it to gmx-users-request at gromacs.org.
> > > > > > > > * Can't post? Read
> > http://www.gromacs.org/Support/Mailing_Lists
> > > > > > > >
> > > > > > > --
> > > > > > > gmx-users mailing list    gmx-users at gromacs.org
> > > > > > > http://lists.gromacs.org/mailman/listinfo/gmx-users
> > > > > > > * Please search the archive at
> > > > > > > http://www.gromacs.org/Support/Mailing_Lists/Search before
> > > posting!
> > > > > > > * Please don't post (un)subscribe requests to the list. Use the
> > > > > > > www interface or send it to gmx-users-request at gromacs.org.
> > > > > > > * Can't post? Read
> http://www.gromacs.org/Support/Mailing_Lists
> > > > > > >
> > > > > > --
> > > > > > gmx-users mailing list    gmx-users at gromacs.org
> > > > > > http://lists.gromacs.org/mailman/listinfo/gmx-users
> > > > > > * Please search the archive at
> > > > > > http://www.gromacs.org/Support/Mailing_Lists/Search before
> > posting!
> > > > > > * Please don't post (un)subscribe requests to the list. Use the
> > > > > > www interface or send it to gmx-users-request at gromacs.org.
> > > > > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > > > > >
> > > > > --
> > > > > gmx-users mailing list    gmx-users at gromacs.org
> > > > > http://lists.gromacs.org/mailman/listinfo/gmx-users
> > > > > * Please search the archive at
> > > > > http://www.gromacs.org/Support/Mailing_Lists/Search before
> posting!
> > > > > * Please don't post (un)subscribe requests to the list. Use the
> > > > > www interface or send it to gmx-users-request at gromacs.org.
> > > > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > > > >
> > > > --
> > > > gmx-users mailing list    gmx-users at gromacs.org
> > > > http://lists.gromacs.org/mailman/listinfo/gmx-users
> > > > * Please search the archive at
> > > > http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> > > > * Please don't post (un)subscribe requests to the list. Use the
> > > > www interface or send it to gmx-users-request at gromacs.org.
> > > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > > >
> > > --
> > > gmx-users mailing list    gmx-users at gromacs.org
> > > http://lists.gromacs.org/mailman/listinfo/gmx-users
> > > * Please search the archive at
> > > http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> > > * Please don't post (un)subscribe requests to the list. Use the
> > > www interface or send it to gmx-users-request at gromacs.org.
> > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > >
> > --
> > gmx-users mailing list    gmx-users at gromacs.org
> > http://lists.gromacs.org/mailman/listinfo/gmx-users
> > * Please search the archive at
> > http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> > * Please don't post (un)subscribe requests to the list. Use the
> > www interface or send it to gmx-users-request at gromacs.org.
> > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >
> --
> gmx-users mailing list    gmx-users at gromacs.org
> http://lists.gromacs.org/mailman/listinfo/gmx-users
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/Search before posting!
> * Please don't post (un)subscribe requests to the list. Use the
> www interface or send it to gmx-users-request at gromacs.org.
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>



-- 
Cordiali saluti, Dr.Oteri Francesco



More information about the gromacs.org_gmx-users mailing list