Previous Thread
Next Thread
Print Thread
white noise in CHARMM force formula
#24078 04/22/10 10:31 AM
Joined: Apr 2005
Posts: 74
G
guan Offline OP
Forum Member
OP Offline
Forum Member
G
Joined: Apr 2005
Posts: 74
Can you please explain/comment on how the white noise was introduced to MD simulations?
Thanks,

G

Re: white noise in CHARMM force formula
guan #24081 04/22/10 04:29 PM
Joined: Sep 2003
Posts: 8,500
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,500
Can you be more specific?


Rick Venable
computational chemist

Re: white noise in CHARMM force formula
rmv #24086 04/22/10 06:38 PM
Joined: Apr 2005
Posts: 74
G
guan Offline OP
Forum Member
OP Offline
Forum Member
G
Joined: Apr 2005
Posts: 74
For example, for the default integrator, does charmm have any random force (white noise) term added in the total force for each atom?

Thanks,

G

Re: white noise in CHARMM force formula
guan #24088 04/22/10 07:20 PM
Joined: Sep 2003
Posts: 8,500
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,500
Langevin dynamics adds a random force, and uses the default integrator (LEAPfrog).


Rick Venable
computational chemist

Re: white noise in CHARMM force formula
rmv #24089 04/22/10 07:31 PM
Joined: Sep 2003
Posts: 4,794
Likes: 2
Forum Member
Offline
Forum Member
Joined: Sep 2003
Posts: 4,794
Likes: 2
Note that "random" is not the same as "white noise"; I would argue that the inclusion of Langevin friction and random forces means that your integrator is different; ie standard Newtonian leap-frog/verlet integration is diffferent from Langevin integration.

The only random numbers used with the default integrator in CHARMM (leap-frog) are for initial velocity assignment and possibly (under user control) for velocity re-assignments during heating/equilibratioon; these are however not white noise.


Lennart Nilsson
Karolinska Institutet
Stockholm, Sweden
Re: white noise in CHARMM force formula
lennart #24094 04/22/10 07:56 PM
Joined: Sep 2003
Posts: 8,500
rmv Online Content
Forum Member
Online Content
Forum Member
Joined: Sep 2003
Posts: 8,500
The algorithm changes with the added forces, but the integration scheme is still coordinate based (not velocity).


Rick Venable
computational chemist


Moderated by  lennart, rmv 

Link Copied to Clipboard
Powered by UBB.threads™ PHP Forum Software 7.7.4
(Release build 20200307)
Responsive Width:

PHP: 5.6.33-0+deb8u1 Page Time: 0.006s Queries: 26 (0.002s) Memory: 0.9307 MB (Peak: 1.0320 MB) Data Comp: Off Server Time: 2020-10-01 19:55:16 UTC
Valid HTML 5 and Valid CSS