From: Josh Vermaas (vermaasj_at_msu.edu)
Date: Mon May 01 2023 - 12:02:21 CDT

Hi Christos,

Yes, wrapping chains won't be required if all you care about is the RMSD
or RMSF or other quantities that are *just* related to the protein.
Wrapping only enters into this when you care about the protein
interacting with its surroundings, since then you'd want to make sure
that the nearest neighbor atoms are accurate. But if all you want to
know is if the protein is stable or not, that is independent of the
environment. Basically, if you look at the trajectory in VMD, and it
looks like there aren't flutters created by the PBC, you'll be fine. You
will likely still need to align your protein though, so that you
actually are calculating the minimum RMSD without a rotation or
translation component.

-Josh

On 5/1/23 12:43 PM, Efthymiou, Christos wrote:
> Hi Josh,
>
> I would like to clarify something. I tried wrapping the whole protein
> complex, one chain individually, and sequentially wrapping each chain
> individually, but still had the same issue. However, if I just used
> the command pbc unwrap -sel "protein", the problem was solved. My
> understanding is that the command removes the periodic boundary
> conditions, allowing the protein to stay together rather than being
> split across the simulation box. Without wrapping the chains after
> unwrapping, the protein can still drift outside of the box and will
> not be in the center of the solution. However, for calculations such
> as RMSD, RMSF, etc. or any calculations only pertaining to the protein
> component of the system, they should be correct without wrapping the
> protein chains, right? Please let me know if this is correct or if it
> is necessary to wrap the chains for all calculations.
>
> Best,
> Christos
> ------------------------------------------------------------------------
> *From:* Vermaas, Josh <vermaasj_at_msu.edu>
> *Sent:* Friday, April 28, 2023 2:20 AM
> *To:* Efthymiou, Christos <christos.dereschuk.20_at_ucl.ac.uk>;
> vmd-l_at_ks.uiuc.edu <vmd-l_at_ks.uiuc.edu>
> *Subject:* Re: vmd-l: Random Sudden Increase in RMSD Near End of
> Simulation
>
> ⚠ Caution: External sender
>
>
> Hi Christos,
>
> If you have a protein complex that is split apart by PBC, you need to
> put it back together again for RMSD analysis. pbc unwrap is supposed
> to do this, but may run into issues if there are any frames that do
> not have periodic box information loaded into them. Trajectory files
> like dcds or xtcs have these by default, but not all inputs do.
> Another thing to try is to PBC wrap around *one* of the protein
> domains. That declares that one domain is the "center", and the
> algorithm will try and put everything within 1 box of that central
> protein. If you select over the whole "protein" selection, the center
> of the selection will be the center of mass of the separated proteins,
> which is probably in solution somewhere, and isn't what you want.
>
> -Josh
>
> On 4/28/23 02:58, Efthymiou, Christos wrote:
>> Hello,
>>
>> I am trying to calculate the RMSD for each frame in my simulation,
>> but I noticed that there are two sections near the end (around
>> 100-200 frames each) where the RMSD suddenly jumps from around 2 to 6
>> angstroms. I have visualized a portion of the trajectory and it
>> appears that the protein-protein complex is partially crossing the
>> periodic boundary condition and causing this issue. In my script, I
>> have used the command pbc wrap -centersel "protein" -compound residue
>> -all prior to calculating the RMSD. I have also tried using pbc
>> unwrap -sel "protein" -all first, but the issue persists. I also
>> tried having the pbc wrap command happen again 25%, 50%, and 75% of
>> the way through the calculation by placing it within the RMSD
>> calculation loop with an interval, but this still did not solve the
>> issue.
>>
>> How can I ensure that my RMSD calculation is corrected, and the
>> protein-protein complex does not cross the periodic boundary
>> condition if pbc wrap is not working?
>>
>> Best,
>> Christos
>
> --
> Josh Vermaas
> vermaasj_at_msu.edu <mailto:vermaasj_at_msu.edu>
> Assistant Professor, Plant Research Laboratory and Biochemistry and Molecular Biology
> Michigan State University
> vermaaslab.github.io

-- 
Josh Vermaas
vermaasj_at_msu.edu
Assistant Professor, Plant Research Laboratory and Biochemistry and Molecular Biology
Michigan State University
vermaaslab.github.io