From: Sachin Natesh (snatesh_at_uchicago.edu)
Date: Tue Aug 16 2016 - 13:01:45 CDT
Hi Josh,
Thanks for your response. The screenshots I provided are from the unwrapped
trajectory. As you suggest, the membrane looks fine when wrapped. The
barostat was set to have constant ratio. Our goal is to study protein-lipid
interaction. For this purpose, does it suffice to unwrap the protein alone
while leaving the lipid wrapped?
Best,
Sachin
On Tue, Aug 16, 2016 at 12:25 PM, Vermaas, Joshua <Joshua.Vermaas_at_nrel.gov>
wrote:
> Hi Sachin,
>
> Do you have wrapall turned on? To me this looks like perfectly normal
> membrane simulation, except that randomly you have net momentum in the top
> and bottom leaflets that aren't equal (which happens alot! System momentum
> being zero doesn't mean the component momenta are). Usually, this is hidden
> by the fact that wrapall on will rewrap lipids as they drift automatically.
> This can also be fixed by postprocessing in vmd (see pbctools and its wrap
> command). In short, your dynamics are probably fine, it just looks a little
> funny when viewed. The one caveat I need to add to my statement is that you
> didn't specify if your barostat is set to have a constant ratio or not.
> Membrane and water hav different compressibility moduli, so you can't use
> the standard isotropic barostat.
>
> Happy Membraning!
> -Josh
>
> On 08/16/2016 11:04 AM, Sachin Natesh wrote:
> Hello,
>
> I am simulating peptides above a lipid bilayer in the NPT ensemble. After
> ~200ns of production, the membrane layers appear to undergo in-plane
> shearing (in the unwrapped trajectory, see links for screenshots). My
> periodic boundary conditions minimize the x/y distance between images, so
> the membrane is 'infinite' in the x-y plane. The only deformation is the
> inter-layer shear and fragmentation at the edges of the membrane (see link
> for membrane w/ multiple periodic images).
>
> Has anyone observed this behavior for similar systems? Might this be an
> artifact of unwrapping, or improper handling of image centering?
>
> Any help is appreciated!
>
> Thanks,
> Sachin Natesh
>
> https://www.dropbox.com/s/i1rlm0zm7zki46l/after.png?dl=0
> https://www.dropbox.com/s/313ywglb0mghlzw/before.png?dl=0
> https://www.dropbox.com/s/h60m05g6akt9n7z/top.png?dl=0
>
>
This archive was generated by hypermail 2.1.6 : Tue Dec 27 2016 - 23:22:22 CST
