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

[XAvier Periole]

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