Re: Boundary potential of colvar alpha not working?

From: Yun Luo (luoyun724_at_gmail.com)
Date: Thu Oct 13 2011 - 09:03:46 CDT

Hi Jerome,

Maybe I asked too many questions at the same time. Sorry about the rush. My first question is: if the wallpotential can be applied to colvar alpha without turning on the extendedLagrangian? Because in ug, it says "System forces are available currently only for the following components: distance, distanceZ, distanceXY, angle, dihedral, rmsd, eigenvector and gyration." . The colvars alpha is not included in calculating system forces.

Thanks!

Lyna
 

On Oct 13, 2011, at 8:54 AM, Jérôme Hénin wrote:

> Hi Lyna,
>
> I can't say anything more specific unless you tell me roughly what you
> are trying to achieve.
>
> Jerome
>
> On 13 October 2011 15:21, Yun Luo <luoyun724_at_gmail.com> wrote:
>> Thanks Jerome! I think I tried the wall potential constant until 1000. But I can try with even higher constant (Is it a quadratic harmonic potential?) But what about the extendedLagrangian force constant? Does it necessary ? If yes, how to balance to wall force constant with extendedlagrangian force constant?
>>
>> Thanks!
>>
>> Lyna
>>
>> On Oct 13, 2011, at 7:14 AM, Jérôme Hénin wrote:
>>
>>> Hi Lyna,
>>>
>>> The "wall" potentials are actually soft, so the colvar is not strictly
>>> constrained to be inside the range. In particular, if the colvar is
>>> subject to high forces, the bias is not always enough to overcome
>>> them. This could be the case here. Since the full range of the
>>> coordinate is [0, 1], a force constant of 100 is not huge. You could
>>> try a higher value.
>>>
>>> Cheers,
>>> Jerome
>>>
>>>
>>> 2011/10/12 Yun Luo <luoyun724_at_gmail.com>:
>>>>
>>>> Dear NAMD supporter,
>>>>
>>>> I'm using "alpha" colvar for a deca-alanine peptide in vacuum. I set up the boundary for alpha 0.0 to 2.5 as below. And I increased the wallconstant and decrease the extendedFluctuation as below. However, during 2 ns the configuration just stays stable at the region outside of by alpha boundary. It seems to me that no matter what constant I use, the boundary potential is not working for alpha colvar, no matter extendedLagrangian is turned on or off. I tested with distance colvar, the boundary potential works well though.
>>>> The outputAppliedForce is fluctuating between -2000 and 2000, but average around 0.
>>>> The outputSystemForce is 0.
>>>> The MISC in NAMD's energy output is fluctuating between 0 and 5, but average around 1.
>>>> The Boundary in NAMD's energy output is 0.
>>>>
>>>> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>>>>
>>>> colvarsTrajFrequency 100
>>>> colvarsRestartFrequency 100000
>>>>
>>>> colvar {
>>>> name alpha
>>>> width 0.01
>>>> lowerboundary 0.0
>>>> upperboundary 0.25
>>>> lowerwallconstant 100.0
>>>> upperwallconstant 100.0
>>>> extendedLagrangian on
>>>> extendedFluctuation 0.0005
>>>> outputValue on
>>>> outputVelocity on
>>>> outputEnergy on
>>>> outputSystemForce on
>>>> outputAppliedForce on
>>>>
>>>> alpha {
>>>> residuerange 1-10
>>>> psfSegID HLX
>>>> hbondcoeff 0.5
>>>> angleref 88
>>>> angletol 15
>>>> hbondcutoff 3.3
>>>> hbondexpnumer 6
>>>> hbondexpdenom 8
>>>> }
>>>> }
>>>>
>>>> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>>>>
>>>> Thanks a lot!
>>>>
>>>> Lyna
>>>>
>>>>
>>
>>

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:20:53 CST