I am doing membrane protein simulation. I plotted the RMSD graph to check the convergence of my protein. I found unusual deviation in RMSD of bacbone of my protein.
below I have attached the RMSD and RMSF image as well as protein structure for your reference.
Red color in protein is a loop from residue 107 to 124. magenta color is an another loop at C terminal from residue 212 to 283.
I also generated the pdb trajectory at zero ns and 11 ns and calculated the rmsd in pymol by aligning the trajectory.
rmsd in pymol = (2.462)
please help me with this issue. I also want to know what is the maximum rmsd acceptable in pymol as well as in simulation result.
Thank you.
--What is the reference structure? It makes sense to use the initial coordinates of the system as the reference coordinates for calculating RMSD.
--MD simulation is divergent: the system is extremely unlikely to ever revisit the initial configuration (coordinates).
--RMSD can fluctuate a lot or a little, depending on the system and parameters of the simulation.
--RMSD can be skewed by few highly mobile elements, such as solvent-exposed loops, and by rigid-body hinge-mediated motion of (sub)domains.
from my understanding, RMSD can be used as indicator of the different conformations of a protein. In this view, simulations should continue until you achieve a constant conformation, i.e. until RMSD is stable in a narrow range. the best way to achieve this in the start of a simulation is to have the protein well optimized (energy well minimized). otherwise the protein will keep deviating until it reaches a stable minimum. Every protein is different and even every system is different, for example I will assume that a protein in a lipid membrane will also be influenced by movements of lipids.
you have to distinguish between the RMSD of the whole molecule and the RMSD of fluctuating fragments. You have to identify rigid and flexible fragments, only then you can understand what stands behind...for other aspects see comment of Martin Klvana
Thank you all for your response. I have added the RMSF plot here to see the fluctuation of each residue. In RMSF plot I found residue 107-124 is fluctuating more, which is a loop region in my protein that can be seen from uploaded pymol image. I also found that there is a very big loop at C terminal from residue 212 to 283. Is this is the reason of getting extremely unlikely RMSD.
Thank you.
Loop regions have higher deviation than normal, and your protein has prominently 2 loop regions. First region 107-124 make sense, but why you have a long C-ter loop? Did you model this protein or it is an x-ray structure?
By aligning in Pymol, you should have seen the deviation in loop region particularly C-ter. Your question states, its a membrane protein, so did you model membrane as well, or protein in water simulation you have performed?
Finally, 20ns is not sufficient time after that convergence can be evaluated. For that, longer (and repetitive) simulations are required.
HTH.
Thanks for your valuable suggestion Muhammad Ayaz Anwar. The protein I have taken is the model protein, which I further model it into POPC bilayer. Actually I didn't get template structure for c-terminal region so I got a big loop while modeling. I am checking the impact of mutation in protein structure and my mutation is lying in between 1-200. Will it be significant to remove these big loop of C -terminal ?
Thank you
It is better to model this region, since, my assumption is that (this protein looks like a channel rather than just TM part of any receptor) it may form another helix in protein that may interact with your desired mutated helix/region.
However, you should model the whole protein through I-tasser, QUARK (it has limit of upto 200 amino acids, you can model in 2 pieces and then assemble).
Finally, it depends, whether this region has any role in protein stability or it may interact with membrane. If this region does not have any role, you can remove it.
Good luck...
Hi,
Can you post the RSMD just for the region of your snapshot? in case that is stable, Folding-unfolding event could also be taking place in the unstructured region. You can dump the structure and check at the three different points of your original RMSD plot to understand better.
Hi Arjun
I calculated the RMSD of region 107-124 as 6.16 by using pymol command (align structure and resi 107-124, structure and resi 107-124 ). likewise when I am calculating the RMSD of C terminal region by using command (align 11 and resi 212-283, zero and resi 212-283) it is not showing the result.
Am I doing correct to calculate rmsd of loop region.
Thank you.
- Badges
- Science topic