From: Peter Freddolino (petefred_at_ks.uiuc.edu)
Date: Wed Jun 27 2007 - 10:35:31 CDT
Hi Alessandro,
the only place where COM motion will show up in NAMD if you remove
initial center of mass motion is from artifacts due to PME, which is
known not to conserve linear momentum perfectly. If you want to
calculate a diffusion coefficient, you may wish to correct for this
artifact by using the zeroMomentum keyword. This is preferrable to
removing COMMotion upon each restart because it modifies the dynamics of
the system in a consistent fashion, rather than removing com drift
abruptly at each restart.
Peter
Alessandro Cembran wrote:
> I think his concern is more than just getting the right rmsd. I had a
> similar concern in the past, when my goal was trying to get an
> estimate of the diffusion coefficient of a protein in a lipid
> membrane. What I found is that the entire membrane together with the
> protein was drifting in a given direction, so to get rid of this bias
> I had to subtract the COM of the entire system (I also found in
> literature that there are people that even remove the COM for each
> layer independently, see Edholm J.C.P. 2001, 115, 4938 or Lyubartsev
> J.P.C.B 2006, 110, 14362). In order to do this one had to write some
> script to unwrap the coordinates from the PBC, if wrapall is specified.
> Furthermore I noticed a weird behavior, in my first simulations I was
> using "ComMotion no", meaning that every time I restarted my
> simulation the total velocity of the system was reset to zero. This
> caused an artificial oscillation to show up in the drifting of the
> system, clearly at the same frequency (a few ns) with which I was
> restarting my simulations. Since then I turned it off and I am
> simulating with "ComMotion yes" and I am removing the total box
> drifting COM afterwards.
> I'd also like to hear some comments about this.
>
> Alessandro
>
> Cesar Luis Avila wrote:
>> To get rid of the diffusion why don't you just align the C alpha
>> atoms of your protein between frames before calculating the rmsd.
>>
>> Ilya Chorny escribió:
>>> Hello All,
>>>
>>> After analyzing my trajectories of a large protein in a membrane, I
>>> noticed the my protein shifts about 3 A after about 10 ns. Would
>>> one expect diffusion on this time scale? Taking a diffusion constant
>>> of about 10^-8 cm^2/s the conversion into MD units is .1A^2/ns (if I
>>> did the math right), thus 3 A seems a bit much. Any thoughts?
>>>
>>> I am also working with a manic cluster which kills my jobs every so
>>> often, at which point I restart the trajectories with new initial
>>> veloocites. Could restarting with new initial velocities cause a
>>> shift/diffusion.
>>>
>>> Finally, I do not zero the CM motion of the system. Does that mean I
>>> have a non-zero CM velovity, and thus I should expect the system to
>>> drift as a whole?
>>>
>>> Thanks,
>>>
>>> Ilya
>>>
>>>
>>> --
>>> Ilya Chorny Ph.D.
>>
>
>
This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:44:54 CST