[gmx-users] REMD average acceptance ratio

XAvier Periole x.periole at rug.nl
Sat May 11 17:10:54 CEST 2013 

You are simulating in vacuo! Otherwise the temperature gaps are way too large … 

If you want to analyse the sampling at 300 K, I would suggest you start you first temperature lower, around 280/285 may be. At least to have your second temperature at 300 K. 

the value of c has absolutely not importance … the temperature distribution has … make some test to see how the acceptance ratio evolves … 

On May 11, 2013, at 5:05 PM, bharat gupta <bharat.85.monu at gmail.com> wrote:

> Dear Sir,
> 
> Here's the temperature range that I got form t-remd :
> 1 300
> 2 323.7
> 3 348.75
> 4 375.23
> 5 403.22
> 6 432.83
> 7 464.14
> 8 497.24
> 9 532.26
> 10 569.32
> 11 608.51
> 
> 
> according the above equation c should be somewhere around 2.37.
> 
> 
> On Sat, May 11, 2013 at 11:10 PM, XAvier Periole <x.periole at rug.nl> wrote:
> 
>> 
>> Well, actually things do not look so good. But is it possible that grace
>> is actually no able to plot things correctly? You have line going
>> throughout the plot from complete-left to complete-right!
>> 
>> I am do not know what the t-rems calculator does but apparently it is not
>> optimal in your case. Did you try to use the simple rule Tn=T0 x exp(n c),
>> where T0 is close to your starting temperature and c is a constant that you
>> can tune and will define the spacing between the temperatures. From your
>> current data you can guess the spacing and thus the c value you need. Note
>> that the exchange ratio is quickly converging in the simulation so you can
>> make a few trials …
>> 
>> On May 11, 2013, at 1:40 PM, bharat gupta <bharat.85.monu at gmail.com>
>> wrote:
>> 
>>> Dear Sir,
>>> 
>>> Thank you for your reply. I choose the temperature distribution using
>>> t-remd calculator. Here's the link for index and temp files .
>>> https://www.dropbox.com/s/uvwsdqjix49lg93/remd_index.png
>>> https://www.dropbox.com/s/78vcnaxzpgeekti/remd_temp.png?m.
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> On Sat, May 11, 2013 at 12:04 AM, XAvier Periole <x.periole at rug.nl>
>> wrote:
>>> 
>>>> 
>>>> The replicas seem indeed to have exchange. Using a colour for the #
>>>> replicas would help.
>>>> 
>>>> I could not access to the first link.
>>>> 
>>>> Note also that the increase of exchange ratio with the temperature
>> suggest
>>>> the distribution of the temperature is not optimal and may be with
>> regular
>>>> intervals? You want to use a exponential distribution.
>>>> 
>>>> On May 10, 2013, at 4:53 PM, bharat gupta <bharat.85.monu at gmail.com>
>>>> wrote:
>>>> 
>>>>> Dear gmx members,
>>>>> 
>>>>> I have posted the same question previously , but I didn't get any
>> reply.
>>>>> So, if anyone can help me out ...
>>>>> 
>>>>> I performed a REMD simulation on a peptide 384 atoms (24 residues). In
>>>>> total 11 replicas were simulated for a period of 50ns each. The
>> exchange
>>>>> was allwoed at every 1000 steps. 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
>>>>> Repl      .16  .16  .16  .17  .18  .21  .24  .26  .28  .30
>>>>> Repl  number of exchanges:
>>>>> Repl     0    1    2    3    4    5    6    7    8    9   10
>>>>> Repl     2038 2007 2065 2117 2182 2587 3022 3213 3554 3703
>>>>> Repl  average number of exchanges:
>>>>> Repl     0    1    2    3    4    5    6    7    8    9   10
>>>>> Repl      .16  .16  .17  .17  .17  .21  .24  .26  .28  .30
>>>>> 
>>>>> The acceptance ratio for each replica and average acceptance ratio is
>> as
>>>>> calculated below :-
>>>>> 
>>>>>              accp. ratio
>>>>> 2038 0.16304
>>>>> 2007 0.16056
>>>>> 2065 0.1652
>>>>> 2117 0.16936
>>>>> 2182 0.17456
>>>>> 2587 0.20696
>>>>> 3022 0.24176
>>>>> 3213 0.25704
>>>>> 3554 0.28432
>>>>> 3703 0.29624
>>>>> 0.211904 (avg accp ratio) (Is this value correct ??)
>>>>> 
>>>>> The Pdes used while generating temp. range was also 0.2. Does that mean
>>>>> that  replicas have exchanged for the given temp.range ??.  Here's the
>>>> link
>>>>> for both remd_temp and remd_index files (
>>>>> https://www.dropbox.com/s/uvwsdqjix49lg93/remd_index.png<
>>>> 
>> https://www.dropbox.com/s/uvwsdqjix49lg93/remd_index.png;cid=1368069857486-810
>>>>> )
>>>>> , (
>>>>> https://www.dropbox.com/s/78vcnaxzpgeekti/remd_temp.png?m<
>>>> 
>> https://www.dropbox.com/s/78vcnaxzpgeekti/remd_temp.png?m&cid=1368069857486-810
>>>>> 
>>>>> )
>>>>> 
>>>>> --
>>>>> 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
>>>> 
>>> 
>>> 
>>> 
>>> --
>>> Bharat
>>> Ph.D. Candidate
>>> Biomolecular Engineering Laboratory
>>> Pusan National University
>>> South Korea
>>> Mobile no. - 010-5818-3680
>>> --
>>> 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