Tutorial-l Mailing List

From: Athreya, Nagendra Bala Murali (nathreya_at_illinois.edu)
Date: Fri Dec 16 2016 - 12:20:15 CST

Hi Jeff,

I followed your instructions. I reduced the size of the system, removed all the gridForces commands in the NAMD file, added rigidBonds, increased the time step to 2 fs.

The minimization runs fine but when I try to run equilibration, it throws this: "ERROR: Constraint failure in RATTLE algorithm for atom 2350!". I tried to increase the margin and then it runs fine. But when I load the dcd file of the equilibration, I see that more than half of the DNA translocates through the pore without any external electric field (within 100,000 steps) and then stops there for many steps (~400,000).
Is there a particular reason for this to happen?
How can I avoid this to happen? Does switching off comMotion help?

Sincerely,
Nagendra Athreya
________________________________________
From: Jeff Comer [jeffcomer_at_gmail.com]
Sent: Monday, December 05, 2016 5:48 PM
To: Athreya, Nagendra Bala Murali
Subject: Re: tutorial-l: DNA not translocating - Nanopore tutorial

> Also, though the model with Graphene and DNA is just around ~10,000 atoms, when I add the water and ions, it increases to ~400,000 atoms. Is there any way to reduce the number of water atoms?

It is always better to start working with small systems to understand
the basic phenomena and use bigger systems only when you need them.
Start with a 10bp molecule and a smaller surface of graphene and you
will have a lot fewer water molecules.

Jeff

末末末末末末末末末末末末末末末末末沫覧覧覧
Jeffrey Comer, PhD
Assistant Professor
Institute of Computational Comparative Medicine
Nanotechnology Innovation Center of Kansas State
Kansas State University
Office: P-213 Mosier Hall
Phone: 785-532-6311
Website: http://jeffcomer.us

On Mon, Dec 5, 2016 at 4:27 PM, Athreya, Nagendra Bala Murali
<nathreya_at_illinois.edu> wrote:
> Hi Jeff,
>
> Thank you very much for your reply. I did go through the paper. I think I kept the gridForce just to make sure that DNA doesn't stick to Graphene.
>
> The dielectric constant is HUGE when it is non-rigid. I did not know that. This is a great info!
>
> I have made the changes and submitted the job for running the simulation. I will get back to you after it has run for few ns. (I am hoping that with these changes I should be able to see the translocation by the first few ns).
>
> Also, though the model with Graphene and DNA is just around ~10,000 atoms, when I add the water and ions, it increases to ~ 400,000 atoms. Is there any way to reduce the number of water atoms?
>
> Thank you once again for your help and time.
>
> Sincerely,
> Nagendra Athreya
> ________________________________________
> From: Jeff Comer [jeffcomer_at_gmail.com]
> Sent: Thursday, December 01, 2016 9:14 AM
> To: Athreya, Nagendra Bala Murali
> Cc: tutorial-l_at_ks.uiuc.edu
> Subject: Re: tutorial-l: DNA not translocating - Nanopore tutorial
>
> Hi,
>
> First, if you're not already aware of it, you should have a look at this paper:
>
> http://dx.doi.org/10.1021/nl301655d
>
> What are you using the gridForce for? I don't see any reason to use it
> with this system. Does the DNA translocate with the gridForce turned
> off?
>
> The DNA tends to adhere to the surface, which may be why you aren't
> seeing translocation. It also likely depends on the size of the pore.
> You might also want to check that lengthZ is being set correctly. Add
> the following to the bottom of your file:
>
> print "lengthZ $lengthZ"
>
> Although we did not use rigidBonds in the tutorial (this should
> probably be amended), as has been noted on this list previously, it's
> an incorrect application of the CHARMM force field. CHARMM is
> parameterized for rigid water and rigid bonds to hydrogen. Also, using
> rigidBonds will make your simulation faster by allowing you to use a 2
> fs timestep. Moreover, flexible TIP3P water has an unrealistically
> large dielectric constant, which is important for simulations such as
> yours involving highly charged species. I calculate a dielectric
> constant of 165 for flexible TIP3P and 99 for rigid TIP3P.
>
> I would recommend this:
> # integration
> rigidBonds all
> timestep 2
> nonBondedFreq 1
> fullElectFrequency 2
> stepsPerCycle 10
>
> Your system is HUGE. Maybe you should try with a small system first
> with a short piece of DNA until you get the protocol established. A
> minimal system could probably have ~40,000 atoms. This might also help
> you to pin down what is causing the lack of translocation without
> having to run for weeks.
>
> Regards,
> Jeff
>
>
> 末末末末末末末末末末末末末末末末末沫覧覧覧
> Jeffrey Comer, PhD
> Assistant Professor
> Institute of Computational Comparative Medicine
> Nanotechnology Innovation Center of Kansas State
> Kansas State University
> Office: P-213 Mosier Hall
> Phone: 785-532-6311
> Website: http://jeffcomer.us
>
>
> On Wed, Nov 30, 2016 at 8:46 PM, Athreya, Nagendra Bala Murali
> <nathreya_at_illinois.edu> wrote:
>> Hello,
>>
>> I am modeling a DNA nanopore system similar to the one mentioned in the nanopore tutorials. I am using Graphene instead of Silicon Nitride.
>>
>> I have a dsDNA of about 60 bp long. The system consists of ~1.5 million atoms (water + ions included).
>>
>> This is NAMD file I have:
>>
>> *******************************************************************************************
>> set voltage 1.0
>> set xsc scaled_${sys}_ions.xsc
>>
>> set this ${sys}_${voltage}V_1
>> #set last ${sys}_eq
>> set next ${sys}_${voltage}V_2
>>
>> numsteps 20000000
>> structure ${sys}_ions.psf
>> coordinates scaled_${sys}_ions.pdb
>>
>> outputName $next
>> binCoordinates $this.restart.coor
>> binVelocities $this.restart.vel
>> extendedSystem $this.restart.xsc
>> #extendedSystem $xsc
>>
>> # temperature control
>> langevin on
>> langevinTemp 295
>> langevinFile ${sys}_langevin.pdb
>> langevinCol B
>>
>> # parameters
>> parameters ../c32b1/toppar/par_all27_prot_lipid.prm
>> parameters ../c32b1/toppar/par_all27_na.prm
>> parameters silicon_nitride.par
>> paraTypeCharmm on
>> exclude scaled1-4
>> 1-4scaling 1
>>
>> switching on
>> switchDist 10
>> cutoff 12
>> pairListDist 14
>> #margin 2.5
>>
>> # integraion
>> firsttimestep 14391000
>> timestep 1
>> nonBondedFreq 2
>> fullElectFrequency 4
>> stepsPerCycle 20
>>
>> # output
>> binaryOutput yes
>> binaryRestart yes
>> wrapAll yes
>> wrapNearest yes
>> comMotion yes
>>
>> outputEnergies 1000
>> outputPressure 1000
>> outputTiming 1000
>> xstFreq 1000
>> dcdFreq 5000
>> restartFreq 5000
>>
>> # electrostatics
>> pme on
>> pmeGridSizeX 192
>> pmeGridSizeY 192
>> pmeGridSizeZ 320
>>
>> # external forces
>> constraints on
>> consKCol B
>> consRef ${sys}_restrain.pdb
>> consKFile ${sys}_restrain.pdb
>>
>> gridforcechecksize off
>> gridforce on
>> gridforceFile specific.pdb
>> gridforceCol B
>> gridforceChargeCol O
>> gridforcePotFile specific2-2.dx
>> gridforceScale 2 2 2
>> gridforceCont1 on
>> gridforceCont2 on
>> gridforceCont3 off
>>
>> set inStream [open $xsc r]
>> set lengthZ [lindex [lindex [split [read $inStream] \n] 2] 9]
>> close $inStream
>> eFieldOn yes
>> eField 0.0 0.0 [expr 23.06054917 * $voltage / $lengthZ]
>>
>> *******************************************************************************************
>>
>> The simulation has run for about 15 ns now and I do not see the DNA translocating in z-direction even a little bit. I had tried the tutorial with same model and parametes and under 20 V before and I could see it run perfectly. I do not ave any clue as to why I can't see the DNA move down the pore.
>>
>> Is there something I am doing wrong? I kindly request someone to look over this and guide me through fixing errors if I have any.
>> **
>> ** Visit http://www.greatcircle.com/majordomo/ for more info on majordomo
>> **

**
** Visit http://www.greatcircle.com/majordomo/ for more info on majordomo
**