[Spam:*****] [gmx-users] REMD analysis

bharat gupta bharat.85.monu at gmail.com
Thu May 16 01:50:01 CEST 2013 

The plots that I showed in my last mail were for all replicas. I tried
plotting the first 500 ps of replica_index and replica_time files. I think
the plots look fine, and there could be problem with the plotting tool .
Here the link for both files ,
https://www.dropbox.com/s/2g16mlxfsme4rx2/replica_temp.bmp
https://www.dropbox.com/s/8jfs0b9whu6j7lo/replica_index.bmp

Now regarding the high acceptance ratio which is 0.5 , I came across a
paper (http://www.pnas.org/content/100/13/7587.full.pdf), here they have
mentioned that their average acceptance ratio ranged between 30 to 80%. I
have a question here, how did they calculate the range for the average
acceptance ratio or is it average ratio for each replica . Actually, this
is the reference I am following. I am also interested in peptide folding
simulation, similar to this article.

I want to know, whether the average acceptance ratio that I have got for my
trial simulation is correct , together with the replica_temp and
replica_remd plots. Can I proceed for large production runs to complete my
experiment ??




On Tue, May 14, 2013 at 6:34 PM, XAvier Periole <x.periole at rug.nl> wrote:

>
> The interval between the exchange trial affect the efficiency of REMD but
> not the the exchange ratio (at least in principle).
>
> In you case I am not sure what the plot are showing! Are these showing all
> the replicas? what are the units?
>
> On May 14, 2013, at 5:07 AM, bharat gupta <bharat.85.monu at gmail.com>
> wrote:
>
> > Dear Sir,
> >
> > Here's the result for the REMD trial with large temperature gaps.
> >
> > Temp. distribution : 280.0 294.9 310.7 327.3 344.7 363.1 382.5 402.9
> 424.4
> > 447.1 471.0 496.1 522.6 550.5 579.9 610.8
> >
> > Out of md16.log :
> >
> > Replica exchange statistics
> > Repl  249 attempts, 125 odd, 124 even
> > Repl  average probabilities:
> > Repl     0    1    2    3    4    5    6    7    8    9   10   11   12
> > 13   14   15
> > Repl      .40  .34  .38  .43  .43  .36  .45  .40  .37  .48  .47  .45  .47
> > .44  .46
> >
> > Repl  number of exchanges:
> > Repl     0    1    2    3    4    5    6    7    8    9   10   11   12
> > 13   14   15
> > Repl       50   42   46   52   57   40   58   49   42   53   61   63   56
> > 57   58
> >
> > Repl  average number of exchanges:
> > Repl     0    1    2    3    4    5    6    7    8    9   10   11   12
> > 13   14   15
> > Repl      .40  .34  .37  .42  .46  .32  .46  .40  .34  .43  .49  .51  .45
> > .46  .46
> > Average acceptance ratio : 0.46
> >
> > But, the repli_index.xvg and replica_temp.xvg files still shows that the
> > replicas does not exchange equally well .
> >
> > https://www.dropbox.com/s/zkbwpuj7l2o282b/replica_index.png
> > https://www.dropbox.com/s/0c8gp584v1hvlbx/replica_temp.png
> >
> > what could be wrong in this case?? Is it the mdp file settings or
> implicit
> > solvent setting. Does the time to replica to exhange also affects their
> > swapping ??
> >
> >
> >
> > On Tue, May 14, 2013 at 12:24 AM, XAvier Periole <x.periole at rug.nl>
> wrote:
> >
> >>
> >> You need to increase the temperature gaps indeed if you want acceptance
> >> ratio ~0.2/0.3. But again this won't work with the water …
> >>
> >> It is not clear what happens in your index file but probably a problem
> >> from grace to plot so many points … you can try to increase the "Max
> >> drawing path length" in the preference menu of grace.
> >>
> >> On May 13, 2013, at 4:22 PM, bharat gupta <bharat.85.monu at gmail.com>
> >> wrote:
> >>
> >>> Dear Sir,
> >>>
> >>> I repeated the simulation again for 25 replicas with the following
> temp.
> >>> distribution .
> >>>
> >>> 280
> >>> 289.1
> >>> 298.5
> >>> 308.2
> >>> 318.2
> >>> 328.6
> >>> 339.3
> >>> 350.3
> >>> 361.7
> >>> 373.5
> >>> 385.6
> >>> 398.1
> >>> 411.1
> >>> 424.4
> >>> 438.3
> >>> 452.5
> >>> 467.2
> >>> 482.4
> >>> 498.1
> >>> 514.3
> >>> 531.0
> >>> 548.3
> >>> 566.1
> >>> 584.5
> >>> 603.5
> >>> 623.2
> >>>
> >>> The output of md.log file is :-
> >>>
> >>> Replica exchange statistics
> >>> Repl  24999 attempts, 12500 odd, 12499 even
> >>> Repl  average probabilities:
> >>> Repl     0    1    2    3    4    5    6    7    8    9   10   11   12
> >>> 13   14   15   16   17   18   19   20   21   22   23   24   25
> >>> Repl      .63  .63  .62  .62  .61  .61  .60  .60  .59  .59  .58  .59
>  .59
> >>> .60  .60  .61  .62  .62  .63  .64  .64  .65  .65  .66  .66
> >>>
> >>> Repl  number of exchanges:
> >>> Repl     0    1    2    3    4    5    6    7    8    9   10   11   12
> >>> 13   14   15   16   17   18   19   20   21   22   23   24   25
> >>> Repl     7822 7752 7816 7760 7639 7628 7511 7442 7375 7332 7312 7424
> 7408
> >>> 7410 7522 7559 7684 7697 7878 7927 7917 8073 8151 8208 8266
> >>>
> >>> Repl  average number of exchanges:
> >>> Repl     0    1    2    3    4    5    6    7    8    9   10   11   12
> >>> 13   14   15   16   17   18   19   20   21   22   23   24   25
> >>> Repl      .63  .62  .63  .62  .61  .61  .60  .60  .59  .59  .58  .59
>  .59
> >>> .59  .60  .60  .61  .62  .63  .63  .63  .65  .65  .66  .66
> >>>
> >>> The average acceptance ration is around 0.6 which is still high.
> >>>
> >>> The link for replica_temp,replica_index :
> >>> https://www.dropbox.com/s/c7soajnwc3uww8j/replica_temp.png
> >>> https://www.dropbox.com/s/wvx82m4c6cnsfit/replica_index.png
> >>>
> >>> The temp files look better but the index file looks weird ...
> >>>
> >>> Do i need to experiment with the gap difference in order to get the
> >>> required ration of 0.2-0.3 ?? There is some problem with the .mdp file
> >>> settings??
> >>>
> >>> --
> >>> Bharat
> >>> --
> >>> 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
>

More information about the gromacs.org_gmx-users mailing list