Re: abrupt deviation in rmsd values

From: Peter Freddolino (petefred_at_ks.uiuc.edu)
Date: Thu May 03 2007 - 19:25:27 CDT

Hi Monika,
as it stands now, you can't dynamically change the langevin bath
temperature. You can accomplish exactly what you want by breaking your
one simulation up into a series of subsequent simulations, each with an
appropriate bath temperature. Otherwise, you won't be able to use
langevin dynamics for the heating, and will have to instead do periodic
velocity reassignment as discussed previously.
Peter

Monika Sharma wrote:
> Hii!!
> Thanks all for the suggestions. But one think I want to ask is that
> how can we change the Langevin dynamics (for temp) in between the run
> commands. Actually I wanted my system to gradually heat up and the
> rmsd changes that occur. And I want to use Langevin dynamics as per
> heating up. But inbetween the run commands Langevin dynamics cannot be
> changed, since the simulation aborts saying "Langevin Parameters
> cannot set off or on". I also tried introducing the LangevinTemp $Temp
> into the loop. I dont know how to fix that prob. The langevin dynamics
> have to be set in the beginning; and cannot be hold inbetween on and
> off as per my experience. (Do correct me if I am wrong or if there is
> any other way??) So the reassigning parameters also doesnt work as I
> want.. :-( Regards,
> Monika
>
> Peter Freddolino wrote:
>
>> Hi Monika,
>> three things:
>> First, it's unclear how you're actually planning on heating the system.
>> There's nothing in your temperature increment loop that is actually
>> changing the system temperature (such as, say temperature $temp).
>> Second, you've set the langevin temperature to your desired temperature
>> from the beginning -- this will mean that your system is always coupled
>> to a 480 K bath, which probably is not what you want. I'd recommend
>> holding off on the langevin dynamics until after your heating is done
>> (you've turned off the Langevin barostat, but not the thermostat).
>> Finally, you may want to turn langevinHydrogen off, since you're using
>> rigidBonds.
>>
>> You should have a look at Temperature rescaling (or reassignment) in the
>> NAMD manual for the type of heating you're looking to do; see
>> http://www.ks.uiuc.edu/Research/namd/2.6/ug/node30.html.
>>
>> Best, Peter
>>
>> Monika Sharma wrote:
>>
>>
>>> Dear All,
>>> I am new to this MD thing and newer to NAMD. I am trying to unfolding
>>> simulation of a small protein. And the protocol I am using is like:
>>>
>>> 1. Energy minimization with fixed atoms, then with no fixed atoms
>>> 2. Heating the system form 0 to 480 k using Langevin Dynamics
>>> 3. Position restratined MD, fixing backbone atoms, and then
>>> 4. free MD equilibration.
>>>
>>> The run ends ok, but when i plot the rmsd over time. the plot shows
>>> the deviation of about 4.0 in the initial phase of 10 ps or so... :-(
>>> When I look at the log file created, its seen that the temperature
>>> rises abruptly from 0 to 167K in a first go; whereas I have given it
>>> the incr of 48K with each run. Dont know where the problem arises.
>>> May be the sudden increase of the temp is destabilizing the protien to
>>> such an extent. Can there be other possible way for getting the
>>> temperature increased gradually?? I am attaching the .conf file and
>>> the rmsd data file for the reference. If anybody can help me out.
>>>
>>> ----------------------------------------------------------------------------------------------------------------------
>>>
>>>
>>> the execution code is like..
>>> #Minimization
>>>
>>> #run to get into scripting mode
>>> minimize 0
>>>
>>> # turn off pressure control
>>> langevinPiston off
>>> #minimize non-backbone atoms
>>> minimize 2500
>>>
>>> #minimize all atoms
>>> fixedAtoms off
>>> minimize 2500
>>>
>>> #Heat with Ps restrained
>>> for {set temp 0} {$temp <= 480} {incr temp 48} {
>>> run 10000
>>> }
>>> #equilibrate volume with Ps restrained
>>> langevinPiston on
>>> run 100000
>>> #equilibrate volume without restraints
>>> constraintScaling 0
>>> run 500000
>>> ------------------------------------------------------------------------------------------------------------------------------
>>>
>>>
>>>
>>>
>>> ------------------------------------------------------------------------
>>>
>>>
>>> #molecules
>>> set import ionized
>>> set minim runmod_incr
>>> set export run_270407 structure ${import}.psf
>>> coordinates ${import}.pdb
>>> bincoordinates ${minim}.coor
>>> extendedSystem ${minim}.xsc
>>>
>>> #forcefield parameters
>>> paratypecharmm on
>>> parameters /home/guest/projects/c32b1/toppar/par_all22_prot.inp
>>> exclude scaled1-4
>>> 1-4scaling 1.0
>>>
>>> #integrator parameters timestep 1.0
>>> fullElectFrequency 4
>>>
>>> outputenergies 100
>>> outputTiming 100
>>> outputPressure 100
>>> #initial config
>>> set init_temp 0
>>> set desired_temp 480
>>> temperature $init_temp
>>>
>>> #outputfiles and restart options
>>> outputname ${export}
>>> binaryoutput no
>>> restartname run_270407_res
>>> restartfreq 1000
>>> binaryrestart yes
>>> dcdfreq 1000
>>> xstfreq 1000
>>>
>>> #others
>>> nonbondedFreq 1
>>> rigidBonds all
>>> rigidTolerance 1.00E-7
>>> stepspercycle 20
>>>
>>>
>>> #simulation spacing partitioning
>>> switching on
>>> switchDist 10
>>> cutoff 12
>>> pairlistdist 16
>>> margin 1
>>>
>>>
>>> #Periodic Boundary Conditions
>>> cellBasisVector1 38 0 0
>>> cellBasisVector2 0 41 0
>>> cellBasisVector3 0 0 45
>>> cellOrigin 4 -4 -16
>>>
>>> wrapWater on
>>> wrapAll on
>>> wrapNearest on
>>>
>>> #PME params
>>> PME yes
>>> PMEGridSpacing 1.0
>>> PMEGridSizeX 45
>>> PMEGridSizeY 45
>>> PMEGridSizeZ 45
>>>
>>> #Fixed atoms for initial heating up steps
>>> fixedAtoms on
>>> fixedAtomsForces on
>>> fixedAtomsFile zwv_fix_backbone.pdb
>>> fixedAtomsCol B
>>>
>>> #Restrained atoms for initial heating-up
>>> constraints on
>>> consRef zwv_constraint_ca.pdb
>>> consKFile zwv_constraint_ca.pdb
>>> consKCol B
>>>
>>> #langevin dynamics params
>>>
>>> #langevin temp control
>>> Langevin on
>>> LangevinDamping 4
>>> LangevinHydrogen on
>>> LangevinTemp $desired_temp
>>> #langevin pressure control
>>> LangevinPiston on
>>> LangevinPistonTarget 1.01325
>>> LangevinPistonPeriod 200
>>> LangevinPistonDecay 100
>>> LangevinPistonTemp $desired_temp
>>> useGroupPressure yes
>>>
>>> #Minimization
>>>
>>> #run 1 step to get into scripting mode
>>> minimize 0
>>>
>>> #turn off pressure control until later
>>> langevinPiston off
>>> #minimize non-backbone atoms
>>> minimize 2500
>>> #minimize all atoms
>>> fixedAtoms off
>>> minimize 2500
>>> #Heat with Ps restrained for {set temp 0} {$temp<=480} {incr temp 48} {
>>> run 10000
>>> }
>>>
>>> #equilibrate volume with Ps restrained
>>> langevinPiston on
>>> run 100000
>>> #equilibrate volume without restraints
>>> constraintScaling 0
>>> run 500000
>>>
>>>
>>>
>>>
>>
>>
>>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:46:15 CST