Re: SMD behavior

From: poker_at_physics.usyd.edu.au
Date: Wed Nov 21 2007 - 18:18:13 CST

Did you check your timesteps? Read the userguide on SMD - the constraint
position is determined based on velocity times the current timestep.
-> You'll need to change the timestep every simulation and/or change the
reference file when you pull.

Quoting Adam Fraser <adam.n.fraser_at_gmail.com>:

> I am trying to push and pull a molecule in successive simulations as is
> done
> in Atomic Force Microscopy (AFM).
>
> I've done this before and never had a problem with it until now and I
> can't
> figure out what's wrong.
>
> *The conditions:*
>
> During each "push" simulation, I use settings:
> structure 2HEX_solv_HEAVY-13-16.psf
> ..
> SMD on
> SMDFile 2HEX_SMD_PUSH.pdb
> SMDk 6.6
> SMDVel 0.000016 #(A/fs) 20A/1200000fs
> SMDDir -1 0 0
>
> During each "pull" simulation, I use settings:
> structure 2HEX_solv_HEAVY-16.psf
> ..
> SMD on
> SMDFile 2HEX_SMD_PULL.pdb
> SMDk 6.6
> SMDVel 0.000016 #(A/fs) 20A/1200000fs
> SMDDir 1 0 0
>
> 2HEX_SMD_PUSH.pdb selects 2 atoms to push whereas 2HEX_SMD_PULL.pdb only
> selects one atom to pull.
>
> 2HEX_solv_HEAVY-13-16.psf specifies that the two pushed atoms are VERY
> heavy
> 2HEX_solv_HEAVY-16.psf specifies that only the single pulled atom is
> VERY
> heavy
> ..all other atoms have normal mass
>
> eg:
> Hexadecane molecule: /\/\/\/\/\/\/\/\
> PUSH: /\/\/\/\/\/\/*\/\* <-- F
> PULL: /\/\/\/\/\/\/\/\* --> F
> *asterisks mark the atoms being pushed/pulled
>
> *What is happening:*
> During the first push simulation ONLY, the molecule is moving in the -x
> direction 20Angstroms nice and smoothly over a period of 1.2ns.
>
> ALL remaining simulations (pull,push,pull,push...) behave unexpectedly.
> The
> molecule moves VERY quickly backward and springs back and forth over
> ~30A
> distance about every 60ps. During this oscillation, the center of the
> spring appears to be drifting in the direction of the applied force.
>
> I repeatedly compared the first push simulation to the others and found
> no
> differences whatsoever that could account for the difference in behavior
> I'm
> seeing.
>
> Does anyone have any ideas or suggestions to help me out with this
> problem?
> I'd greatly appreciate any help.
>

----------------------------------------------------------------
This message was sent using IMP, the Internet Messaging Program.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:45:35 CST