From: Jérôme Hénin (
Date: Mon Jan 28 2019 - 08:03:37 CST


You can restrain the center of mass of the protein laterally using a
hamonic restraint on a distanceXY coordinate, between a subset of atoms of
the protein (alpha carbons?) and a dummy atom.


On Sat, 26 Jan 2019 at 16:49, jeevan gc <> wrote:

> Hi Josh,
> Thank you for your valuable suggestions.
> I completely agree with you that the protein should be allowed to move in
> the membrane system. As you suggested, I decided to use selectConstraints to
> only use lateral constraints. However, for the multimeric proteins like
> dimer, tetramer, pentamer etc., the drift caused some problems, especially
> if I want to count hydration of pore (for example in calcium channels
> tetrameric proteins) along the membrane z-axis.
> These problems could be solved by pbcwrap after simulations as well.
> Do you recommend using a higher damping coefficient (increase to 2 or 3
> ps-1) instead of constraints?
> Thank you.
> Jeevan
> On Fri, Jan 25, 2019 at 9:59 AM Vermaas, Joshua <>
> wrote:
>> If you want to disallow lateral protein motion completely, why not just
>> use constraints?
>> Just
>> tag the protein and use selectConstraints to only use lateral constraints.
>> However, in practice you really shouldn't care if your membrane drifts, and
>> my usual practice has been to ignore it unless I need to make an animation.
>> Proteins and membranes move, and that's just a fact of life. I'd
>> recommend turning off wrapAll, and then just rewrapping afterward in VMD,
>> possibly even recentering the protein prior to rewrapping.
>> -Josh
>> On 2019-01-24 16:00:34-07:00 wrote:
>> Dear all,
>> We have come across an issue on protein and membrane regular MD
>> simulations. We are also aware of the fact that this issue is just about
>> visualization and should be taken care by the Periodic Boundary
>> Conditions. The membrane size is sufficient enough to avoid any overlap in
>> periodic image as the protein drift from the edge of the membrane. The wrap
>> all is also turned on. The namd mailing list provided us important
>> information and suggestions about this issue but still not able to
>> completely solve this issue.
>> Below are list of simulations we would like to practice to avoid the
>> drift of the protein from the membrane edge.
>> 1. Increased the step6.6 equlibration simulation step obtained from
>> CHARMM-GUI to 10ns or higher.
>> 2. Increase the damping constant from 1 to 2 ps.
>> 3. Increased the size of the membrane.
>> 4. Applied the minimal restraint to the protein and the membrane system.
>> 5. NPT/NVT minimization
>> We tried Simulation #1, it didn't help. The protein drifted from the edge
>> of the membrane. Is it necessary to increase damping coefficient to 2ps
>> as suggested by James Starlight?
>> If these only visualization issues, increasing the size of the membrane
>> will increase the simulation time and we don't benefit much.
>> Simulation #4, restraining either protein of membrane COM may create an
>> artifact in protein membrane interactions.
>> As CHARMM-GUI minimization steps suggested, we performed step6.1 and
>> step6.2 in NVT, step6.3 to step6.6 in NPT ensemble. Production run was
>> performed on NPT.
>> Are there any best practices to avoid such drift of protein in the
>> membrane? Any advice will be greatly appreciated.
>> Thank you in advance.
>> Jeevan
>> --
>> *Jeevan B. GC, Ph.D*
>> *Post Doctoral Research Associate*
>> *Department of Pharmaceutical Sciences*
>> *Washington State University*
>> *Spokane, WA 99224 , **USA*
> --
> *Jeevan B. GC, Ph.D*
> *Post Doctoral Research Associate*
> *Department of Pharmaceutical Sciences*
> *Washington State University*
> *Spokane, WA 99224 , **USA*

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2020 - 23:17:09 CST