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

[bharat gupta]

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
>