Re: Nanotube Simulation

From: Axel Kohlmeyer (akohlmey_at_gmail.com)
Date: Mon Apr 14 2014 - 13:53:53 CDT

On Apr 14, 2014 7:55 PM, "Kenno Vanommeslaeghe" <
kenno_at_outerbanks.umaryland.edu> wrote:
>
> Yeah, we CHARMM users have been using our benzene atom type as well for
casual studies involving graphenes and nanotubes. The L-J are fairly
transferable and the charges cannot reasonably be set to a value far from 0
for a lot of these cases, so we've been working on the assumption that "it
cannot be too bad". I'd be more comfortable with more validation, but this
is easier said than done.

Yeah, one has to be content with understanding, that this model is more an
impenetrable hydrophobic cylindrical void than a CNT and thus be careful
not to overinterpret any observations. For polar biopolymers in aqueous
solution that is a fair model, I think; beyond that (vacuum, nonpolar
solvents) are likely dragons to be found.

> Also, we have a biomedical focus and are not experts in these manybody
and Tersoff potentials. This is something that probably needs to be further
looked into from our side.

Manybody potentials are only important, if you want to bend, twist and
strain the CNT. At that point, I would be much more concerned, since that
will impact the electronic structure locally and then the assumption that
polarization and other Coulombic effects can be on average neglected (by
setting the atomic charges to zero) is no longer good.

>
> On 04/11/2014 06:41 PM, Axel Kohlmeyer wrote:
>>
>> On Fri, Apr 11, 2014 at 6:18 PM, Kenno Vanommeslaeghe
>> <kvanomme_at_rx.umaryland.edu> wrote:
>>>
>>> There's one thing bugging me a bit in this discussion... The first
sentence
>>> in Miro's original post was: "I would like to run a NAMD calculation of
a
>>> nanotube interacting with a protein via non-bonded interactions using
CHARMM
>>> FF." I'm not familiar with the tools that were mentioned in the
subsequent
>>> answers, and would like to know what force field / atom type is
assigned to
>>> the nanotube. If it is an unofficial extension of the CHARMM FF, what
is the
>>
>>
>> the plugin only builds the geometry and the bond topology. most people
>> that i know are using the plugin (which is likely a very small subset)
>> then convert the geometry to a LAMMPS topology and use a manybody
>> potential to describe the CNT or graphene sheet. some dummy values are
>> assigned to make sure VMD writes out a valid file.
>>
>>> citation? And on what base is it thought to interact correctly with a
CHARMM
>>> protein?
>>
>>
>> the studies that i know about, use the atom type for a benzene carbon
>> (but they don't use CHARMM, but Amber (throughout)) and make the atoms
>> representing the CNT immobile or use a manybody potential like REBO or
>> AIREBO to describe the CNT structure itself. codes like LAMMPS allow
>> this kind of overlaying different potential types within certain
>> boundaries. sometimes also a Tersoff potential used.
>>
>> axel.
>>
>>>
>>>
>>>
>>> On 04/11/2014 02:33 PM, MEHRAN MB wrote:
>>>>
>>>>
>>>> Hi Miro,
>>>> I guess, you need to select your nano tube atoms then write pdb and psf
>>>> file.
>>>>
>>>> like :
>>>> set mol [atomselect top "all resname CNT"]
>>>> $mol writepdb 1.pdb
>>>> $mol writepsf 1.psf
>>>>
>>>> best,
>>>> Mehran
>>>> Uottawa
>>>>
>>>>
>>>> On Fri, Apr 11, 2014 at 1:13 PM, Miro Hodak <mhodak127_at_gmail.com
>>>> <mailto:mhodak127_at_gmail.com>> wrote:
>>>>
>>>> Maybe this is too obvious, but I do not see it - and I have
experience
>>>> with both NAMD and VMD.
>>>>
>>>> I can save coordinates (pdb or other formats) once a nanotube is
>>>> created, but I see no way of getting a PSF file. I looked into the
>>>> User's Guide and tried the Automated PSF builder plugin, but I
cannot
>>>> figure this out.
>>>>
>>>> If you can be more specific ...
>>>>
>>>> Thanks again,
>>>> Miro
>>>>
>>>>
>>>> On Fri, Apr 11, 2014 at 1:05 PM, Axel Kohlmeyer <
akohlmey_at_gmail.com
>>>> <mailto:akohlmey_at_gmail.com>> wrote:
>>>>
>>>>
>>>> On Apr 11, 2014 6:56 PM, "Miro Hodak" <mhodak127_at_gmail.com
>>>> <mailto:mhodak127_at_gmail.com>> wrote:
>>>> >
>>>> > I know of that one, that is how I built my nanotube, but I
do
>>>> not see a way to get a PSF file from this plugin once the
nanotube
>>>> is built.
>>>>
>>>> How to save files is explained in the vmd user's guide.
>>>>
>>>> >
>>>> > Thanks again,
>>>> > Miro
>>>> >
>>>> >
>>>> > On Fri, Apr 11, 2014 at 4:36 AM, Axel Kohlmeyer
>>>> <akohlmey_at_gmail.com <mailto:akohlmey_at_gmail.com>> wrote:
>>>> >>
>>>> >> On Thu, Apr 10, 2014 at 7:02 PM, Miro Hodak
>>>> <mhodak127_at_gmail.com <mailto:mhodak127_at_gmail.com>> wrote:
>>>> >> > Dear all,
>>>> >> >
>>>> >> > I would like to run a NAMD calculation of a nanotube
>>>> interacting with a
>>>> >> > protein
>>>> >> > via non-bonded interactions using CHARMM FF. This looks
>>>> quite similar to
>>>> >> > what is
>>>> >> > done in a tutorial on water flow through nanotubes.
However,
>>>> that tutorial
>>>> >> > already provides a prebuilt PSF file for the nanotube
used
>>>> there. I want to
>>>> >> > work
>>>> >> > with a different nanotube (20,0) and looking at the
tutorial
>>>> and other
>>>> >> > resources
>>>> >> > I cannot figure out how to create a PSF file for an
>>>> arbitrary nanotube I may
>>>> >> > want to work with.
>>>> >> >
>>>> >> > Can anybody help?
>>>> >>
>>>> >> http://www.ks.uiuc.edu/Research/vmd/plugins/nanotube/
>>>> >>
>>>> >> >
>>>> >> > Thank you,
>>>> >> > Miro
>>>> >>
>>>> >>
>>>> >>
>>>> >> --
>>>> >> Dr. Axel Kohlmeyer akohlmey_at_gmail.com
>>>> <mailto:akohlmey_at_gmail.com> http://goo.gl/1wk0
>>>>
>>>> >> College of Science & Technology, Temple University,
>>>> Philadelphia PA, USA
>>>> >> International Centre for Theoretical Physics, Trieste.
Italy.
>>>> >
>>>> >
>>>>
>>>>
>>>>
>>>
>>
>>
>>
>

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:22:20 CST