Re: Colvars Orientation Question

From: Jérôme Hénin (jerome.henin_at_ibpc.fr)
Date: Tue Jul 09 2019 - 04:49:07 CDT

Hi Alexander,

To get a full view of the process, you can enable outputEnergy for the
harmonic bias and outputAppliedForce for the collective variables. If the
deviations are large you should see high energies and forces.

Best,
Jerome

On Fri, 5 Jul 2019 at 21:23, Xander Gonzalez <xandergonz1_at_gmail.com> wrote:

> Thank you for the response. I have checked the colvars.traj file and the
> beginning quaternions are significantly different than the center colvars
> set in the configuration file, and the final quaternions are different than
> the target centers. I also visualized the rotations with the script you
> sent, it seems like the initial rotation is caused by the second collective
> variable rotating about 90 degrees. I have tried setting the target center
> for the second colvar quaternion to be the same as it's center to see if
> this would keep this rotation from happening, but I get the same result.
>
> Best,
> Alexander
>
> On Wed, Jul 3, 2019 at 3:54 AM Jérôme Hénin <jerome.henin_at_ibpc.fr> wrote:
>
>> One last point: quaternions values are not very intuitive. If it helps,
>> you can visualize them as a rotation axis and lines indicating the value of
>> the angle. Load a trajectory in VMD, source this script and use the
>> "start_rotation_display" command, with the name of the orientation colvar.
>>
>> https://github.com/Colvars/colvars/blob/master/vmd/scripts/colvar_display.tcl
>>
>> Best,
>> Jerome
>>
>> On Wed, 3 Jul 2019 at 10:49, Jérôme Hénin <jerome.henin_at_ibpc.fr> wrote:
>>
>>> Hi Alexander,
>>>
>>> the explanation to this mystery might be in the colvars.traj file. Look
>>> in particular for the initial and final values of the orientation
>>> quaternion.
>>>
>>> Jerome
>>>
>>> On Tue, 2 Jul 2019 at 23:50, Xander Gonzalez <xandergonz1_at_gmail.com>
>>> wrote:
>>>
>>>> Thank you for the feedback, I have been trying to run a static harmonic
>>>> restraint and running into some trouble. I have tried setting targetCenters
>>>> to the same as Centers as well as not using targetCenters at all. Both
>>>> result in the protein making what seems to be a complete rotation. Setting
>>>> forceConstant as 1 seems to result in a static protein, but I am not sure
>>>> if this is correct. Did you have any specific ideas about how to create a
>>>> static harmonic restraint? Or do the results that I have been getting imply
>>>> something about the system?
>>>>
>>>> Thank you again,
>>>> Alexander
>>>>
>>>> On Mon, Jul 1, 2019 at 7:02 PM Giacomo Fiorin <giacomo.fiorin_at_gmail.com>
>>>> wrote:
>>>>
>>>>> Hi Alex, it is likely that the combined effect of the two rotations
>>>>> that you are applying is inducing a torque on the entire protein. Can you
>>>>> try first to see if a static harmonic restraint is stable for this choice
>>>>> of variable before starting to move it?
>>>>>
>>>>> Giacomo
>>>>>
>>>>> On Mon, Jul 1, 2019 at 4:01 PM Alexander Gonzalez <
>>>>> acgonzalez_at_uchicago.edu> wrote:
>>>>>
>>>>>> Hello,
>>>>>>
>>>>>> I've been trying to run a simulation on a new structure with
>>>>>> orientation collective variables to separate two portions of a protein
>>>>>> (resid 43-515 and resid 542-1011), I followed the "Exploring Complex
>>>>>> Conformational Transition Pathways" tutorial, using the provided
>>>>>> orientation.tcl to find the orientation quaternions. I changed the
>>>>>> colvars.conf file as is shown below:
>>>>>>
>>>>>> #############################################################
>>>>>> ## Collective Variables ##
>>>>>> #############################################################
>>>>>>
>>>>>> # Global parameters
>>>>>> colvarsTrajFrequency 500
>>>>>>
>>>>>> colvar {
>>>>>> name d1_2
>>>>>> orientation {
>>>>>> atoms {
>>>>>> psfSegID AP1
>>>>>> atomNameResidueRange { CA 43-515 }
>>>>>> }
>>>>>> refPositionsFile CC-complete-model-ChainA-solvate-centered-mini-LF.pdb
>>>>>> }
>>>>>> }
>>>>>>
>>>>>> colvar {
>>>>>> name d3_4
>>>>>> orientation {
>>>>>> atoms {
>>>>>> psfSegID AP1
>>>>>> atomNameResidueRange { CA 542-1011 }
>>>>>> }
>>>>>> refPositionsFile CC-complete-model-ChainA-solvate-centered-mini-LF.pdb
>>>>>> }
>>>>>> }
>>>>>>
>>>>>> harmonic {
>>>>>> name harm
>>>>>> colvars { d1_2 d3_4 }
>>>>>> # initial values
>>>>>> centers { ( 1 , 0 , 0 , 0 ) ( 1 , 0 , 0 , 0 ) }
>>>>>> # target values
>>>>>> targetCenters { ( 0.984807753012208 , -0.009832450473359431 ,
>>>>>> 0.04197699686335437 , 0.16821101136975466 ) ( 0.984807753012208 ,
>>>>>> 0.009832450473359431 , -0.04197699686335437 , -0.16821101136975466 ) }
>>>>>> # force constant in kcal/(mol*rad^2)
>>>>>> forceConstant 10000
>>>>>> # steering protocol time
>>>>>> targetNumSteps 200000
>>>>>> # calculate work
>>>>>> outputAccumulatedWork on
>>>>>> }
>>>>>>
>>>>>> The resulting simulation seems to separate the two protein sections,
>>>>>> but not before rotating and distorting the protein for the first few
>>>>>> frames. I also tried accomplishing this separation with the spin colvars,
>>>>>> where I used the axis.tcl script to get the principal axes. Drawing these
>>>>>> axes seems to show that they are accurate for the rotation that I want to
>>>>>> accomplish. Despite this, the spin colvar simulation also leads to the
>>>>>> structure rotating and distorting in unwanted ways for the first few frames
>>>>>> of the simulation before separating as expected. I would really appreciate
>>>>>> advice for possible things to try regarding this, also let me know if you
>>>>>> want more information about the specific structure/simulation I am
>>>>>> attempting. Thanks for your time!
>>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> Giacomo Fiorin
>>>>> Associate Professor of Research, Temple University, Philadelphia, PA
>>>>> Research collaborator, National Institutes of Health, Bethesda, MD
>>>>> http://goo.gl/Q3TBQU
>>>>> https://github.com/giacomofiorin
>>>>>
>>>>

This archive was generated by hypermail 2.1.6 : Mon Sep 16 2019 - 23:20:42 CDT