VMD-L Mailing List

From: I. Camps (icamps_at_gmail.com)
Date: Sat Aug 31 2019 - 21:32:53 CDT

Hi Joshua,

How do I select the frame to do the alignment? Is it in the "Frame ref:"
field (the same used for RMSD)? If yes, I did it and did not get any
similar to Maestro.

[]'s,

Camps

On Thu, Aug 29, 2019 at 10:25 PM Vermaas, Joshua <Joshua.Vermaas_at_nrel.gov>
wrote:

> Ok, next question, what did you align against? The
> "RMSD_after_alignment.png" looks ok, assuming that you aligned against the
> *last* frame instead of the *first* frame like Maestro did. RMSD measures
> against a specific reference state, and the implementations should give the
> same result, you just need to make sure that the reference state and the
> comparison states are equivalent.
>
> -Josh
>
>
>
> On 2019-08-29 18:13:05-06:00 I. Camps wrote:
>
> The values are very different. I am attaching here a PNG image with both
> curves.
> I am not aware about the Maestro methodology used to calculate RMSD, but,
> thinking in this possibility, I did an alignment in VMD. The figure is also
> attached. The result is very different too.
> Camps
>
> On Thu, Aug 29, 2019 at 7:40 PM Vermaas, Joshua <Joshua.Vermaas_at_nrel.gov>
> wrote:
>
>> How different is different? Changes in the 5th or 6th digit after the
>> decimal point may be due to accumulated round-off errors based on the
>> different order of additions or algorithmic implementation differences. If
>> its larger than that, I suspect you are measuring two different quantities.
>> VMD's RMSD plugins depend on the alignment to be done first, whereas
>> Maestro *might* (?) use the QCP algorithm, which doesn't need an alignment.
>>
>> -Josh
>>
>> PS I can't recall any image or other attachments coming through in the
>> past, so I'm guessing no?
>>
>>
>> On 2019-08-29 12:37:21-06:00 owner-vmd-l_at_ks.uiuc.edu wrote:
>>
>> Dear all,
>> I am newbie using VMD.
>> My first task was to calculate the RMSD from a Molecular Dynamics done
>> using DESMOND.
>> I obtained the C-alpha and ligand RMSD using Maestro interface and using
>> the plug-in withing VMD.
>> The problem I am facing is that the results are different.
>> In the case of C-alpha, in both programs (Maestro and VMD), the selection
>> was: "backbone and name CA".
>> In the case of ligand, in Maestro it is just to select ligand. In VMD,
>> the ligand is identify as "UNK" residue.
>> Also, in both cases, the reference frame is the frame "0".
>> What I am doing wrong?
>> Regards,
>>
>> Camps
>> PS: I sent a previous email to this list with a figure attached. Does
>> this list accept attachments?
>>
>>