Re: SMD constant velocity pulling

From: Gianluca Interlandi (gianluca_at_u.washington.edu)
Date: Thu Dec 02 2010 - 01:38:59 CST

Hi Bidisha,

You use "rigid bonds" in conjunction with multistepping. Although this
might not be wrong, I suspect that in conjunction with pulling it could
lead to temperature increase. I use "rigid bonds" but no multistepping.
Somewhere else it was shown that multistepping alone leads to small
temperature increase over the course of a long simulation. But I'm not
sure whether this is the cause.

Why is "switching" off?

Also, I pull at lower speeds, 1 or 5 A/ns, with a softer spring: 2
kcal/mol. With these parameters I have never needed temp control.
However, these are the parameters which worked for me and I'm not saying
that you should use the same.

Also, I would check that during the 100 ps of equilibration the
temperature has converged. I usually run 200 ps (but again that's a
personal choice).

A tip: you are using PMEGridSizeX 131 and PMEGridSizeY 118
You should use only numbers which contain powers of 2, 3, 5 or 7 in their
prime factor decomposition (e.g., 135 and 120). The runs will be much
faster (and then you can use a lower pulling speed) :)

Gianluca

On Thu, 2 Dec 2010, B. Nandy wrote:

> Hi Gianluca and Jerome,
>
> Thank you for the reply and the discussion. I have an equilibrated system
> consisting of two proteins. I got the equilibrated structure using AMBER
> MD . And after then using AMBER forcefield I ran NAMD CPT for 100 ps
> before starting SMD. I don't think volume is a problem, because I have
> taken a very big water box.
>
> I haven't tried Lowe-Andersen thermostat yet, But before doing that I
> want to check what I have done is correct. Following is my script for SMD
> please have a look at it and let me know if there is any mistake. And
> without using thermostat the rise of temperature depends on the duration
> of SMD run as well as the velocity of pulling.
>
>
> #############################################################
> ## ADJUSTABLE PARAMETERS ##
> #############################################################
>
> # Continuing a job from the restart files
> if {1} {
> set inputname test_gp-cd_npt
> binCoordinates $inputname.restart.coor
> binVelocities $inputname.restart.vel ;# remove the "temperature"
> entry if you use this!
> extendedSystem $inputname.xsc
> }
>
> firsttimestep 0
>
>
> #############################################################
> ## SIMULATION PARAMETERS ##
> #############################################################
>
> # Input
> amber on # Specify this is AMBER force field
> parmfile gp120-cd4-v2.top # Input PARM file
> #ambercoor md_prod_900_ps.restart # Input coordinate file
> ambercoor md_prod_900_ps.restart.img.restart # imaged
> outputname test_gp-cd-af_npt-cns-t-smd # Prefix of output files
>
>
> # NOTE: Do not set the initial velocity temperature if you
> # have also specified a .vel restart file!
> #temperature $temperature
>
>
> wrapWater on
> wrapAll on
>
>
> # Force-Field Parameters
> exclude scaled1-4
> 1-4scaling 0.833333 # =1/1.2, default is 1.0
> cutoff 9.
> switching off
> #switching on
> switchdist 10.
> pairlistdist 13.5
>
>
> # Integrator Parameters
> timestep 2.0 ;# 2fs/step
> rigidBonds all ;# needed for 2fs steps
> nonbondedFreq 1
> fullElectFrequency 2
> stepspercycle 10
>
>
> #PME (for full-system periodic electrostatics)
> if {1} {
> PME yes
> cellBasisVector1 130.4269071 0. 0.
> cellBasisVector2 0. 117.7932744 0.
> cellBasisVector3 0. 0 149.4518528
> PMEGridSizeX 131
> PMEGridSizeY 118
> PMEGridSizeZ 150
> }
>
>
> # Constant Temperature Control
> langevin on ;# do langevin dynamics
> langevinDamping 5 ;# damping coefficient (gamma) of 5/ps
> langevinTemp $temperature
> langevinHydrogen no ;# don't couple langevin bath to hydrogens
>
>
> restartfreq 500 ;# 500steps = every 1ps
> dcdfreq 1000
> xstFreq 500
> outputEnergies 100
> outputPressure 100
>
>
> # Fixed Atoms Constraint (set PDB beta-column to 1)
> if {1} {
> fixedAtoms on
> #fixedAtomsFile 900_all_vmd.pdb.ref
> fixedAtomsFile vmd-test-gp-cd-npt.ref.pdb
> fixedAtomsCol B
> }
>
>
> #############################################################
> ## EXTRA PARAMETERS ##
> #############################################################
>
> # Put here any custom parameters that are specific to
> # this job (e.g., SMD, TclForces, etc...)
>
> SMD on
> SMDFile vmd-test-gp-cd-npt.ref.pdb
> SMDk 5
> SMDVel 0.000032 #0 .016A/ps
> SMDDir -0.26928 0.17691 0.94667
> SMDOutputFreq 20
>
> #############################################################
> ## EXECUTION SCRIPT ##
> #############################################################
>
> run 1000000 ;# 2000 ps
>
>
> ----------------------------------------------------------------------------
>
> Thanking you.
> Bidisha
> IISc Bangalore
> India
>
>
>> Thanks Jérôme,
>>
>> I also prefer Langevin to Berendsen. I might give Lowe-Andersen thermostat
>> a try in the future. For now, I continue using Langevin thermo+barostat
>> (except SMD where I use none). Or, do you think I should immediately
>> switch to Lowe-Andersen since it is the best out there?
>>
>> Gianluca
>>
>> On Wed, 1 Dec 2010, Jérôme Hénin wrote:
>>
>>> Hi Gianluca,
>>>
>>> On 1 December 2010 19:49, Gianluca Interlandi
>>> <gianluca_at_u.washington.edu> wrote:
>>>> Hi Jérôme,
>>>>
>>>> Sorry, I have another question. What kind of pressure control is
>>>> recommended
>>>> to use in conjunction with the Lowe-Andersen thermostat?
>>>
>>> In general, I don't know of any requirements. In the particular case
>>> of NAMD, since there are only two choices of barostat and one of them
>>> is Berendsen, to me there is really only one choice. I admit that this
>>> is rooted in heavy personal bias - and an irrational taste for proper
>>> statistical ensembles. Add to that guilt by association with the
>>> Berendsen thermostat, and its sinister retinue of flying ice cubes.
>>>
>>> Cheers,
>>> Jerome
>>>
>>>> On Wed, 1 Dec 2010, Jérôme Hénin wrote:
>>>>
>>>>> Hi Bidisha,
>>>>>
>>>>> This sounds like a very good application for the recently implemented
>>>>> Lowe-Andersen thermostat. It should add virtually no drag.
>>>>>
>>>>> Best,
>>>>> Jerome
>>>>>
>>>>> On 1 December 2010 10:51, B. Nandy <nandy_at_physics.iisc.ernet.in>
>>>>> wrote:
>>>>>>
>>>>>> Dear NAMD user,
>>>>>>
>>>>>> I have  question regarding SMD simulation using constant velocity
>>>>>> pulling.
>>>>>> I have noticed  in different examples that one should switch of  the
>>>>>> constant temperature control during constant velocity SMD pulling.
>>>>>> But
>>>>>> following this method the temperature has increased unreasonably in
>>>>>> my
>>>>>> system.  On the other hand  using langevin on  temperature remains
>>>>>> constant but  I am getting a huge amount of dragging force.  So
>>>>>> please
>>>>>> suggest me how to deal with the dragging force when thermostat is on.
>>>>>>
>>>>>>
>>>>>> Thanking you
>>>>>>
>>>>>> Bidisha Nandy
>>>>>> IISc Bangalore
>>>>>> India
>>>>>>
>>>>>>
>>>>>> --
>>>>>> This message has been scanned for viruses and
>>>>>> dangerous content by MailScanner, and is
>>>>>> believed to be clean.
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>> -----------------------------------------------------
>>>> Gianluca Interlandi, PhD gianluca_at_u.washington.edu
>>>>                    +1 (206) 685 4435
>>>>                    http://artemide.bioeng.washington.edu/
>>>>
>>>> Postdoc at the Department of Bioengineering
>>>> at the University of Washington, Seattle WA U.S.A.
>>>> -----------------------------------------------------
>>>
>>
>> -----------------------------------------------------
>> Gianluca Interlandi, PhD gianluca_at_u.washington.edu
>> +1 (206) 685 4435
>> http://artemide.bioeng.washington.edu/
>>
>> Postdoc at the Department of Bioengineering
>> at the University of Washington, Seattle WA U.S.A.
>> -----------------------------------------------------
>> --
>> This message has been scanned for viruses and
>> dangerous content by MailScanner, and is
>> believed to be clean.
>>
>>
>
>
> --
> This message has been scanned for viruses and
> dangerous content by MailScanner, and is
> believed to be clean.
>
>

-----------------------------------------------------
Gianluca Interlandi, PhD gianluca_at_u.washington.edu
                     +1 (206) 685 4435
                     http://artemide.bioeng.washington.edu/

Postdoc at the Department of Bioengineering
at the University of Washington, Seattle WA U.S.A.
-----------------------------------------------------

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 05:23:26 CST