Re: Re: Re: Large energy fluctuation

From: Marcos Sotomayor (sotomayo_at_ks.uiuc.edu)
Date: Thu Jun 01 2006 - 13:02:55 CDT

Hi Sting,

you are using a multiple time stepping algorithm! electrostatic
interactions are computed every four timesteps (8fs), nonbonded
interactions every two timesteps (4fs), and bonded interactions every
timestep (2fs). Be careful with how you choose your parameters, you
may want to have the "rigid all" option activated and use a 2,1,2 scheme
for your time steps. Check the users guide for more details:
http://www.ks.uiuc.edu/Research/namd/2.6b1/ug/node25.html#SECTION00083700000000000000

Marcos

On Wed, 1 Jun 2005, Sting wrote:

> Hi Marcos,
> I didn't use multiple time stepping algorithm,but I didn't use uniform time step either:
> timestep 2
> nonbondedFreq 2
> fullElectFrequency 4
> stepspercycle 20
> Is that to say that only the energy at a time step that is multiple of (4/2)make sense?
>
>
>
> ======= 2006-06-01 10:48:37=======
>
>> Have a look at
>> http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l/3407.html
>>
>> likely your are using a multiple time stepping algorithm, and that's why
>> you see changes in the electrostatic energy.
>>
>> Marcos
>>
>> On Wed, 1 Jun 2005, Sting wrote:
>>
>>> Hi Cesar,
>> > Thank you for suggestion.This system is a super-helix consisted of
>> three chains,about 110 X 3 residues.I have performed a 10,000 steps
>> minimization and 500ps MD(with CA fixed),then another 500ps free MD was
>> performed.While minimization,the energy fluctuation(sorry,I shuld correct
>> myself,that the changing energy is not VDW but Elect energy) didn't
>> occur,but both the constrain MD and free MD presented the energy shift as
>> said before.
>>> By the way,during solvating the protein,we set the waters departing 3.3 Å from protein (defaut is 2.4 Å),and then did NTV MD. So that, at some specific positions near the protein, huge water holes occured.Did the Elect Energy shift has any relationship with this?
>>> Thank you for your time.
>>>
>>> Sting
>>>
>>>
>>> ======= 2006-06-01 02:12:52 :=======
>>>
>>>> May be you should better equilibrate the system before running MD.
>>>> Regards
>>>> Cesar
>>>>
>>>> Sting wrote:
>>>>
>>>>> Dear all
>>>>> When I performed a MD run with NAMD,every things went well,but when analyzing the energy of the system,I noticed that there was a large fluctuation between every two sequential steps,something like: step VDW energy
>>>>> 259970 -226049.7557
>>>>> 259980 -38243.0418
>>>>> 259990 -226001.2423
>>>>> 260000 -38106.4633
>>>>> I wonder what I should understand this.
>>>>> Any suggestion will be appreciated.
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>         Sting
>>>>>         stg1979_at_emails.bjut.edu.cn
>>>>>           2005-04-08
>>>>>
>>>>>
>>>
>>>
>>>
>
> = = = = = = = = = = = = = = = = = = = =
>
>
>         致
> 礼!
>
>
>         Sting
>         stg1979_at_emails.bjut.edu.cn
>           2005-06-01
>
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:42:07 CST