[gmx-users] Re: Nanotube and DNA with periodic boundary conditions

[Bob Johnson]

> Hello everyone,
> I'm currently studying the interaction between an infinite carbon nanotube 
> and a
> single stranded DNA molecule. I am using the Amber force field for the DNA
> molecule. I am building my system in XLeap and then converting the 
> resulting
> coordinate and topology file to the GROMACS format. However, the periodic
> boundary conditions (to make the nanotube infinite) don't seem to work 
> with
> this topology (despite using pbc = full). It seems that to simulate an 
> infinite
> nanotube I must generate its topology with x2top and use the -pbc option. 
> ONLY
> when I generate the topology this way do I get a stable, infinite 
> nanotube.
> Assuming that this is true, this is how I've been proceeding.
>
> 1. Build system in Amber's XLeap
>
> 2. Convert coordinate and topology file to GROMACS format. This gives the 
> files
> gromos.g96 and gromacs.top
>
> 3. Copy the nanotube coordinates in gromos.g96 to a new file called 
> cnt.g96
>
> 4. Generate the nanotube topology. This is the command I use:
> x2top -f cnt.g96 -o cnt.top -pbc
>
> 5. Overwrite the old nanotube topology in gromacs.top with cnt.top. The 
> topology
> for the DNA, water, and ions remains unchanged.
>
> 6. Edit the box size in gromos.g96 such that the z-dimension of the box is 
> equal
> to the length of the nanotube plus a C-C bond length (so the images match 
> up)
>
> 7. Run the simulation with pbc = full
>
> The result is a mess. Pieces of the nanotube jump to the next image box 
> and
> cause the program to crash. I then tried giving the nanotube carbon atoms 
> a
> huge mass (so they wouldn't move). This prevented the jumping. However, 
> the DNA
> molecule was unstable (pieces moving around frantically) and some of its 
> bonds
> became too large and the simulation crashed.
>
> Simulations of a single infinite nanotube generated from x2top -pbc 
> produce a
> stable structure. Also, simulations of the DNA generated from Amber 
> produce
> stable results. Only when I combine the two systems do I have problems. Is
> there some incompatibility between the x2top topology and the topology
> converted from Amber when I use pbc = full?
>
> It seems that the problem is more than just an energetically unfavorable
> configuration. I've tried many things to try to reduce that: making the
> nanotube carbons more massive, running minimizations, reducing the 
> temperature,
> constraining the bonds. All of these provide either crashes or unstable 
> results.
>
> Does anybody know what the x2top -pbc option does and how it works with
> pbc=full? If I knew that maybe I would have some insight into what's going 
> on.
> I'm really out of ideas here. If anyone could give me some suggestions I 
> would
> really appreciate it!
> Thanks for you time,
> Bob Johnson
>
>
> ------------------------------
>
> Message: 4
> Date: Sun, 14 Aug 2005 09:20:25 +0200
> From: David <spoel at xray.bmc.uu.se>
> Subject: [gmx-users] [Fwd: Position restraints in GROMACS]
> To: gmx-users at gromacs.org
> Message-ID: <1124004025.4484.4.camel at localhost.localdomain>
> Content-Type: text/plain
>
> -------- Forwarded Message --------
> From: Bob Johnson <robertjo at physics.upenn.edu>
> To: spoel at xray.bmc.uu.se
> Subject: Position restraints in GROMACS
> Date: Sat, 13 Aug 2005 16:11:38 -0400
> Hello Dr. Spoel,
> First off, I wanted to thank you for all your responses to my questions on 
> the
> GROMACS mailing list. Secondly, I had a question about position restraints 
> in
> GROMACS. You mentioned that I had to download the CVS version of GROMACS 
> in
> order to use them. I downloaded it but am having problems installing it. 
> Is the
> CVS version really needed? Another person in my group believed that she 
> used
> them on the normal version of GROMACS. I am placing the include posre.itp
> statement at the end of the topology file. I want to restrain atoms 1-2080 
> so I
> placed the following statement:
> [ position_restraints ]
> ;atom   type     fx     fy    fz
> 1        1       1e9   1e9    1e9
> 2        1       1e9   1e9    1e9
> and so on
>
> I am getting the following error message by grompp:
>
> WARNING 1 [file "posre.itp", line 5]:
>  No default Position Rest. types for perturbed atoms, using normal values
>
> and similarly for line 6-10. Would the CVS version alleviate these issues?
> Thanks for your time,
> Bob Johnson
> University of Pennsylvania
> -- 
> David.
> ________________________________________________________________________
> David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group,
> Dept. of Cell and Molecular Biology, Uppsala University.
> Husargatan 3, Box 596,          75124 Uppsala, Sweden
> phone:  46 18 471 4205          fax: 46 18 511 755
> spoel at xray.bmc.uu.se    spoel at gromacs.org   http://xray.bmc.uu.se/~spoel
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>
>
>
> ------------------------------
>
> Message: 5
> Date: Sun, 14 Aug 2005 09:28:37 +0200
> From: David <spoel at xray.bmc.uu.se>
> Subject: Re: [gmx-users] Nanotube and DNA with periodic boundary
> conditions
> To: Discussion list for GROMACS users <gmx-users at gromacs.org>
> Message-ID: <1124004518.4484.9.camel at localhost.localdomain>
> Content-Type: text/plain
>
> On Sat, 2005-08-13 at 19:51 -0400, Bob Johnson wrote:
>> Hello everyone,
>> I'm currently studying the interaction between an infinite carbon 
>> nanotube and a
>> single stranded DNA molecule. I am using the Amber force field for the 
>> DNA
>> molecule. I am building my system in XLeap and then converting the 
>> resulting
>> coordinate and topology file to the GROMACS format. However, the periodic
>> boundary conditions (to make the nanotube infinite) don't seem to work 
>> with
>> this topology (despite using pbc = full). It seems that to simulate an 
>> infinite
>> nanotube I must generate its topology with x2top and use the -pbc option. 
>> ONLY
>> when I generate the topology this way do I get a stable, infinite 
>> nanotube.
>> Assuming that this is true, this is how I've been proceeding.
>>
> What I'd do is the following.
> Build the nanotube topology using x2top, and run it in vacuo in a large
> box to ensure it is stable. Build the DNA topology however you want,
> merge the two topologies using a text editor and merge the coordinates
> too. This is the crucial step. Take care that there is plenty of space
> between DNA and nanotube. Do energy minimization and short posres
> simulation with force constant 1000 (not 1e9 as this will cause high
> frequency vibration which will make the system unstable).
>
> Then add water using genbox and do posres simulation where you allow the
> box axes perpendicular to the nanotube to fluctuate but not the one in
> the direction of the tube.
>
> Then you should be ready for production.
> -- 
> David.
> ________________________________________________________________________
> David van der Spoel, PhD, Assoc. Prof., Molecular Biophysics group,
> Dept. of Cell and Molecular Biology, Uppsala University.
> Husargatan 3, Box 596,          75124 Uppsala, Sweden
> phone:  46 18 471 4205          fax: 46 18 511 755
> spoel at xray.bmc.uu.se    spoel at gromacs.org   http://xray.bmc.uu.se/~spoel
> ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>

Do I need the CVS version to use position restraints. It doesn't seem like 
this feature works. I get
WARNING 1 [file "posre.itp", line 5]:
  No default Position Rest. types for perturbed atoms, using normal values
Thanks,
Bob Johnson