Re: How to "twist" a molecule? (resubmit)

From: Margaret.Cheung_at_mail.uh.edu
Date: Tue Jun 12 2007 - 11:19:36 CDT

Hi,
You are right. All flags for constraints need to be set in order to make it work.
I can see the info of my rotation axis in the log file...
Sincerely,
Margaret

----- Original Message -----
From: Victor Ovchinnikov <ovchinnv_at_mit.edu>
Date: Monday, June 11, 2007 9:57 pm
Subject: Re: namd-l: How to "twist" a molecule? (resubmit)
To: mscheung_at_uh.edu, namd-l_at_ks.uiuc.edu

> Hi, Margaret.
>
> Just a thought.  In the file SimParameters.C, the following
> code turns
> all constraints off if 'constraints' is undefined:
> Do you have "constraints on" elsewhere in your input file?
>
> Victor
>
> ////////////////////
>  if (!opts.defined("constraints"))
>    {
>      constraintExp =
> 0;    
>      constraintScaling =
> 1.0;    
>
>      //****** BEGIN selective restraints
> (X,Y,Z) changes
>      selectConstraintsOn = FALSE;
>      //****** END selective restraints
> (X,Y,Z) changes
>  
>      //****** BEGIN moving constraints
> changes
>      movingConstraintsOn = FALSE;
>      //****** END moving constraints changes
>      //****** BEGIN rotating constraints
> changes
>      rotConstraintsOn = FALSE;
>     //****** END rotating constraints changes
>    }
>    //****** BEGIN rotating constraints changes
>    else {
>      if (rotConstraintsOn) {
>        rotConsAxis = rotConsAxis.unit();
>      }
>    }
> .
> .
> .
> //////////////////////////////////////////////////////
>
> On Mon, 2007-06-11 at 18:49 -0400, Margaret S. Cheung wrote:
> > Hi All,
> >
> > I think I will need to revisit this problem, because I just
> realized
> > that NAMD program is not rotating the molecule even though the
> > configuration file was set so.
> >
> > My configuration file
> >
> > rotConstraints    on
> > rotConsAxis     x y z
> > rotConsPivot   x y z
> > rotConsVel    xx
> >
> > But in the log file, it gave me warning as
> >
> > Warning: The following variables were set in the configuration
> file but
> > were not needed.
> > Warning: rotConstraints
> > ....
> >
> > And I tried to see if rotConsPivot is in the log file, and
> they are not!
> > Does it mean that my NAMD program is not rotating the molecule?
> >
> > Please help!
> > Sincerely,
> > Margaret
> >
> >
> > Grzegorz Jezierski wrote:
> > > Dear Margaret,
> > > Now I think I understand your problem.
> > > As mentioned on p.71 of NAMD manual, Rotating Constraints
> are set up
> > > the way ordinary (i.e. positional) restraints are.
> > > In my example, reference positions for Rotating Constraints
> are
> > > specified in R6G_6CAA.pdb file, and force constants are
> specified in
> > > fixed.pdb file, in 10th column.
> > > A line in my fixed.pdb file which specifies the force
> constant for
> > > Rotating Constraints looks like this:
> > > ATOM     46 O1  
> MOL     2      -
> 0.300  -1.495   2.119  0.00  1.00
> > > If I put e.g. 0.00 instead of 1.00, the Rotating Constraints
> option
> > > for this atom will be switched off.
> > > Now in order to combine his with harmonic restraints,
> include your
> > > desired restraint definition in a separate file (pmf.conf in
> my
> > > example); description on p.82, use 'mcti' keyword. Specify
> > > 'freeEnergy  on'. I used this option for constraining a
> system, not
> > > just for free energy calculations.
> > > Alternatively, you may want to specify 'fixed atoms'. This
> seems like
> > > a more suitable option for your study. In this case, you
> should use a
> > > different column (e.g. 'O' i.e. 9th) in the pdb file.
> > > Hope this helps
> > >
> > > Margaret S. Cheung wrote:
> > >> Hello,
> > >>
> > >> I tried with both rotational and harmonic contraints, but
> now I see
> > >> the possible problem JC mentioned.
> > >>
> > >> The question is "how to set reference positions for
> rotating
> > >> constraints"? The manual simply said that the rotational
> constraints
> > >> parameters are specified in the same manner as for usual
> > >> (static)harmonic constraints". I assume it means that we
> need to set
> > >> consref, conskfile, conskcol for the harmonic constraints.
> > >>
> > >> But if I now included rotational constraints to harmonic
> one, how can
> > >> I set the reference positions for rotation since all it has
> is
> > >> rotConsAxis, rotConsPivot and rotConsVel?
> > >> When I set constraints on, did it mean "all atoms" in the
> system will
> > >> be rotated? Can I apply rotation (twist) to only a part of
> the
> > >> molecules (e.g. excluding water molecules)?
> > >> Thank you.
> > >> Sincerely,
> > >> Margaret
> > >>
> > >> Grzegorz Jezierski wrote:
> > >>> I've run simulations with almost any combination of
> Rotating
> > >>> Constraints, Harmonic Constraints, Fixed Atoms and PMF
> constraints.
> > >>> Here is a relevant part of one of my input files:
> > >>>
> > >>>
> Constraints             on
> > >>>
> consRef                 R6G_6CAA.pdb
> > >>>
> consKFile               fixed.pdb
> > >>>
> consKCol                B
> > >>>
> > >>>
> rotConstraints          on
> > >>>
> rotConsAxis             3.50 4.15 -20.45
> > >>>
> rotConsPivot            3.50 4.15 -24.37
> > >>>
> rotConsVel             0.0030
> > >>>
> freeEnergy              on
> > >>> freeEnergyConfig       
> pmf.conf> >>>
> > >>>
> > >>> Regards,
> > >>> Greg
> > >>>
> > >>>
> > >>> On Sun, 25 Mar 2007, JC Gumbart wrote:
> > >>>
> > >>> 
> > >>>> Your suggestion definitely sounds like it might work
> (using
> > >>>> fixedAtoms that
> > >>>> is); you can't use harmonic constraints and rotating
> constraints at
> > >>>> the same
> > >>>> time though, if I understand the manual correctly. 
> I haven't seen
> > >>>> rotatingConstraints used by anyone myself, so you may
> just need to
> > >>>> experiment
> > >>>> with it.  You could also try using IMD to control
> the forces
> > >>>> interactively.
> > >>>> Let us know how it goes.
> > >>>>
> > >>>>
> > >>>> On Mar 23, 2007, at 11:05 AM, Margaret S. Cheung wrote:
> > >>>>
> > >>>>  
> > >>>>> (I apologize if you are reading this message again. The
> previous
> > >>>>> one perhaps
> > >>>>> was
> > >>>>> bounced because I haven't registered my current email
> address to
> > >>>>> NAMD-L, as
> > >>>>> warned
> > >>>>> by NAMD-L regulators)
> > >>>>>
> > >>>>> Hi NAMD users and developers,
> > >>>>>
> > >>>>> I'd like to "twist" a molecule by applying a torque
> along an axis.
> > >>>>> I am
> > >>>>> looking at "rotating constraints" in the user guide. But
> a simple
> > >>>>> rotation
> > >>>>> is not what I want. Somehow I would want to "fix" a few
> distant
> > >>>>> atoms away
> > >>>>> from the pivot point so that the molecules can be
> twisted
> > >>>>> eventually (ex:
> > >>>>> like winding up a helical structure). If so, would you
> recommend
> > >>>>> including
> > >>>>> "harmonic constraints" or simply assigning a few "fixed"
> atoms to the
> > >>>>> reference pdb file in order to accomplish that? Please
> advise and
> > >>>>> thank you
> > >>>>> for your time.
> > >>>>>
> > >>>>> Margaret S. Cheung
> > >>>>> Assistant Professor
> > >>>>> Department of Physics
> > >>>>> 629C Science and Research 1
> > >>>>> University of Houston
> > >>>>> Houston, TX 77204-5005
> > >>>>> email: mscheung_at_uh.edu
> > >>>>> (O)713-743-8358 (F)713-743-3589
> > >>>>> http://thoth.phys.uh.edu/~mscheung/
> > >>>>>      
> > >>>>    
> > >>>
> > >>>  
> > >>
> > >
> > >
> >
>
>

Margaret S. Cheung
Assistant Professor
Department of Physics
University of Houston
Houston, TX, 77204-5005
Tel: 713-743-8353
http://thoth.phys.uh.edu/~mscheung/

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:44:50 CST