AMBER Archive (2006)

Subject: AMBER: Re: Simulated annealing

From: Carlos Simmerling (carlos_at_csb.sunysb.edu)
Date: Thu Aug 10 2006 - 06:40:24 CDT

in order to change the temperature you want to change TEMP0, not TAUTP.
this will work for either ntt=1 or ntt=3.
just because energy minimization completed does not mean that
a particular GB model will give the correct results for your simulated
annealing. I strongly encourage you to read what others have found
using different solvent models or you will not be able to interpret your
results.

Ruchi Sachdeva wrote:

>Thanks Sir for your reply
>
>I think I was not able to express my problem to u correctly so I got one
>more doubt that I would like to discuss with you. I would have to use TAUTP in
>order to set weight change information with &wt namelist as I want to heat up
>the system to high temperatures and then cool it down. That's why I am
>confused for the value of ntt either 1 or 3 ?
>
>Initially I used igb=1 for carrying out energy minimization prior to SA.
>But system got highly unstable and energy values came out to be NaN. So I
>switched to igb=2 & then energy minimization completed successfully.
>
>Please correct me if I am wrong anywhere.
>
>Thanks again
>
>On Thu, 10 Aug 2006, Carlos Simmerling wrote:
>
>
>
>>you should use ntt=3 and set the gamma_ln variable to a value other
>>than 0. What you should use for this depends very much on what you
>>want to do- for annealing/structure prediction it may be a good idea
>>to use a small value like 1.0, but you should read some articles about
>>what others have used and make the choice for yourself.
>>tautp isn't used in this case.
>>
>>make sure to apply the bugfixes for amber8, I think there was one for
>>doing annealing with ntt=3.
>>
>>also my personal experience with igb=2 has not been good, make sure to
>>read articles that successfully use the GB model for the same kind of
>>research that you are doing. igb=5 works better in my experience but
>>also has some drawbacks. you should keep these in mind when analyzing
>>your results.
>>
>>Ruchi Sachdeva wrote:
>>
>>
>>
>>>Dear Sir
>>>
>>>I want to run simulated annealing algorithm with Generalized Born model
>>>version2 (igb=2). It is mentioned in AMBER8 manual that temperature
>>>coupling options other than ntt=1 should be used for Generalized Born
>>>simulations. ntt=3 is recommended for gb simulations. But on the other
>>>hand, coupling parameter, TAUTP is used when ntt=1 is used for simulated
>>>annealing.
>>>
>>>So please tell me which thermostat i should use for this case and give me
>>>some guidance on how to set values for TAUTP to perform simulated
>>>annealing.
>>>
>>>Thank you very much in advance for your kind help
>>>
>>>Regards
>>>
>>>Ruchi
>>>
>>>
>>>
>>>

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu