From: Monika Madhavi (
Date: Fri Dec 15 2017 - 16:35:22 CST

Hi Ashar,

I did all this analysis using VMD visualizer tool. I loaded my reference
molecule in the Reference option and then loaded the two trajectories in
the Molecule option. After selecting one trajectory, I pressed Align button
first and then RMSD and plotted it. The same thing was done for the second
molecule. So in my understanding, the calculation must be done as you

Trajectory - A(RMSD): {0-r, 1A-r, 2A-r, 3A-r, ... }
Trajectory - B(RMSD): {0-r, 1B-r, 2B-r, 3B-r, ... }

But my 0-r for two molecules are different even if they start from the same
initial configuration.

One point occurred to me is that, I put the atomselection as Backbone in
the calculation. Is it possible for VMD to calculate the alignment for the
entire molecule while RMSD is only calculated for the backbone.

As I don't have my data with me at the moment, I can't tell you the exact
values of two RMSD's. But they were like 0.3 for one and 0.45 for the other.

Thank you.

On Fri, Dec 15, 2017 at 10:58 PM, Ashar Malik <> wrote:

> Ok.
> So this means that the reference is a separate configuration *"r" * which
> means we can say the following:
> Trajectory - A(RMSD): {0-r, 1A-r, 2A-r, 3A-r, ... }
> Trajectory - B(RMSD): {0-r, 1B-r, 2B-r, 3B-r, ... }
> And you are saying that 0-r in trajectory A is not the same as 0-r in
> trajectory B.
> This brings me to the alignment point that I briefly touched earlier.
> How do you align your trajectories. Are trajectory A and trajectory B
> aligned to their respective frame *"0" *or do you align each trajectory
> to an average structure in it or do you align them to this "reference
> external structure" you use to compute the RMSD?
> reference structure to frame 0 should return the same RMSD value if frame
> 0 has not moved. If you align each of the separate trajectories to an
> average frame within it, then frame 0 would have moved in space and you
> should not get the same number between two trajectories.
> Also may I ask, what is the RMSD for
> your reference configuration and frame 0 of one trajectory
> your reference configuration and frame 0 of the other trajectory
> On Sat, Dec 16, 2017 at 1:30 AM, Monika Madhavi <>
> wrote:
>> The reference frame I use is a separate pdb file of the same system. I
>> used the same reference for the calculation of rmsd for both simulations.
>> Given that the starting configuration of both simulations are same and the
>> reference is the same, I expected the RMSD plot to start from the same
>> point. Because my reference is different from the zero time configuration
>> of both the simulations, neither RMSD values would be zero but I expect
>> them to be some single value.
>> Best,
>> Monika
>> On 15 Dec 2017 22:10, "Ashar Malik" <> wrote:
>>> For simplicity I am going to break this problem down.
>>> Your starting frame for each of the two trajectories is the same. Lets
>>> call this *"0"*.
>>> Then your two trajectories can be represented by:
>>> Trajectory - A: {0, 1A, 2A, 3A, ... }
>>> Trajectory - B: {0, 1B, 2B, 3B, ... }
>>> Lets say that trajectory *"A"* is with the constraints and *"B"*
>>> without the constraints.
>>> RMSD is calculated by comparing each frame in a trajectory to a
>>> reference frame. If you didn't change this, the default should have been
>>> frame *"0" *which is the starting frame.
>>> So the following comparisons will have been made.
>>> Trajectory - 1(RMSD): {0-0, 1A-0, 2A-0, 3A-0, ... }
>>> Trajectory - 2(RMSD): {0-0, 1B-0, 2B-0, 3B-0, ... }
>>> If I understand correctly your RMSD plots start at non-zero values.
>>> Right?
>>> Here are some situations.
>>> 1. Do your plots start at frame 0? If they do it must be a value of
>>> 0 for both plots. In which case the starting values should be identical for
>>> both plots because frame 0 is compared to itself in both plots.
>>> 2. Do your plots start at frame 1? In which case why would want them
>>> to be the same? Frame 1A and 1B are not the same. So why should you expect
>>> the same RMSD? (This is what is likely happening).
>>> There may be other reasons - e.g. your trajectory alignment method which
>>> may also result in a value mismatch, but I think the previous two points
>>> are important. If you provide more evidence and that doesn't explain the
>>> difference you can write back and some one will help.
>>> On Fri, Dec 15, 2017 at 11:55 PM, Monika Madhavi <
>>>> wrote:
>>>> Hi all,
>>>> I minimised and equilibrated a protein water system. Then I used the
>>>> final configuration of this system to start two simulations with the same
>>>> parameters except the constraints on backbone atoms. When I calculated the
>>>> rmsd using Rmsd visualiser tool, the rmsd starts from two values for the
>>>> two trajectories. I can't figure out what could be the reason for this. Any
>>>> insight would be appreciated.
>>>> Thank you
>>>> Best,
>>>> Monika
>>> --
>>> Best,
>>> /A
> --
> Best,
> /A

W.A.Monika Madhavi
Lecturer (Probation),
Department of Physics,
University of Colombo.