[gmx-users] GROMACS and CNT/graphene (probably for the 1000th time)

[Alex]

Hey Justin,

Great you had a moment to look at my post.
This is an impossible task without the coordinates input file, because this is now
the realm of solid-state physics, and there really isn't a single
"residue" that can be created once and for all. It could be really
cumbersome to post everything, and sort of useless to everyone at this
point, before we figure this out.

Would you be OK with me emailing you the input along with the simple series of steps? This way you won't
need to mess with the tutorials, I'll just show you exactly what works
on my end and what doesn't.

Thanks,

Alex



JL> On 3/22/15 12:50 PM, Alex wrote:
>> Hi All,
>>
>> I'm sure you've seen this many times before, but I think it might help
>> those who have never done this before, are new to Gromacs, or, like
>> me, haven't done anything relatively advanced with it in a while.
>>
>> So, here goes. My intermediate goal here was a solvated sheet of
>> graphene, so I followed http://www.gromacs.org/Documentation/How-tos/Carbon_Nanotube
>> under "Andrea Minoia's tutorial" pretty much to the word (there is a link
>> with the exact file content). Everything prior to solvating the system
>> worked without a hitch. To solvate the system, I used genbox, making
>> it also update the topology file (created with g_x2top). This is when
>> grompp started screaming about "defaults" duplicates, and, as I tried
>> removing the #include to cnt_oplsaa.ff/forcefield.itp from the
>> topology file and just put all parameters explicitly in there, there was a bunch of other problems.
>> Without going into detail of the latter, I don't think I know what I am doing at this
>> point.
>>
>> Can you please help with this task in a simple & robust step-by-step way?
>> In fact, I'd like to write up an updated tutorial
>> based on our discussion, because, among other things,
>> Andrea's parameters are taken from oplsaa on aromatics, which can and
>> should be further refined. All input is greatly appreciated and I'd
>> especially love to hear from the Gromacs superhero Justin Lemkul. :)
>>

JL> At a conference, so time is limited.  Provide every step of what you did and any
JL> problem you encountered.  Generating the topology is the hard part; if that was
JL> generated successfully, there (in principle) shouldn't be issues. I don't do
JL> CNT simulations and don't follow those tutorials so I don't know how well they
JL> are expected to work.  The .n2t file for such species should be trivial to
JL> create, and if it's derived from OPLS-AA, everything should be very straightforward.

JL> -Justin