David Case points out that this was already found and corrected
in bugfix 25 - if this is applied the annealing should work fine.
Carlos Simmerling wrote:
> Pascal,
> I finally got time to check the code and you're right, there is a
> variable used
> for temperature in Langevin dynamics that does not get updated during
> the MD run.
> This means that weight changes to TEMP0 have no effect for ntt=3.
> I tested a fix and it works, we'll get a bugfix done as soon as I can
> generate it. I'll mail you a modified sander.f if you want (off the
> list to save
> everyone's mailbox).
> Carlos
> pascal.baillod.epfl.ch wrote:
>
>> Hello,
>>
>> Could anybody tell me how temeprature control is performed, when using a
>> Langevin thermostat in amber8? I suppose that it is implemented as
>> described in
>> van Gunstern et al., 1981, if I understand the manual correctly. The
>> temperature
>> control would in this case be derived from the random collision
>> force, that is
>> taken from a temperature dependant gaussian distribution.
>> I suppose simulated annealing could be performed by slowly varying the
>> temperature of the gaussian distribution yielding the random
>> collision force?
>> But I think that is not implemented, as my simulated annealing input
>> file, that
>> works fine for Berendsen thermostat md, doesn't work for a langevin
>> thermostat md.
>>
>> Thank you very much in advance!
>>
>> Pascal
>>
>>
>>
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Tue May 10 2005 - 13:53:00 PDT
