Re: Flat-bottom harmonic position restraints using colvars

From: David Hardy (dhardy_at_ks.uiuc.edu)
Date: Tue Feb 11 2020 - 15:57:40 CST

Hi Daipayan,

Sorry, I just now saw this exchange.

The harmonic wall potential was added in December. You can get it by downloading the nightly build. In the included user documentation, the section discussing extra bonds explains the use of this "wall" potential.

Best regards,
Dave

--
David J. Hardy, Ph.D.
Beckman Institute
University of Illinois at Urbana-Champaign
405 N. Mathews Ave., Urbana, IL 61801
dhardy_at_ks.uiuc.edu, http://www.ks.uiuc.edu/~dhardy/
> On Feb 10, 2020, at 8:41 PM, Daipayan Sarkar <sdaipayan_at_gmail.com> wrote:
> 
> Hi Giacomo, 
> 
> Sorry for the late reply. Thanks for your explanation.  I have modified the range, such that all values are positive. 
> Second, I performed two separate simulations, keeping one harmonic wall restraint on at a time, and it works. 
> Finally, for my original question, I did not get any feedback on how to implement the harmonic wall potential with extrabonds (based on the link you shared).  
> 
> Best Regards,
> Daipayan
> 
> 
> 
> On Mon, Feb 10, 2020 at 7:14 AM Giacomo Fiorin <giacomo.fiorin_at_gmail.com <mailto:giacomo.fiorin_at_gmail.com>> wrote:
> Hi Daipayan, first of all there is no need to specify the lowerWalls (or lowerWallConstant, for that matter) if these are well below the possible range of values: in this case, a distance variable can never go below -2 Å.
> 
> Secondly, the MISC column reports values of the restraint energy for all variables combined (plus, many other kinds of restraints than Colvars ones, if you have them).  So for a frame where d1 is within range, you could still get a positive MISC energy from d2.  Looking at some of the highest points in your plots, this seems to be the case.
> 
> Lastly, your original question was about applying flat-bottom restraints on individual atoms, which is a bit different with what you are doing here.
> 
> Did you get any feedback regarding the new per-atom flat-bottom restraint feature?  The NAMD development team is copied also.
> 
> Giacomo
> 
> On Sun, Feb 9, 2020 at 1:07 PM Daipayan Sarkar <sdaipayan_at_gmail.com <mailto:sdaipayan_at_gmail.com>> wrote:
> 
> Hello, 
> This is a follow-up on our recent discussion for the flat-bottom harmonic potential implemented as part of collective variables. To validate, if the potential is indeed flat-bottom harmonic as documented: (https://colvars.github.io/colvars-refman-namd/colvars-refman-namd.html#sec:colvarbias_harmonic_walls <https://colvars.github.io/colvars-refman-namd/colvars-refman-namd.html#sec:colvarbias_harmonic_walls>), I applied this to the tma-aco example with the following protocol.
> 
> The harmonic walls potential are applied between dummyatoms (x,y,z coordinates of the center of mass) and their respective molecules, tma and aco. The two distances d1 (dummyatom and tma) and d2 (dummyatom and aco) written to the colvars.traj file are plotted against the MISC term in the log file (the two plots are attached). Each point in the scatter plot is a value from the trajectory every 1000 steps. The result shows that the MISC potential is flat-bottom (in-shape), however, there are points within the cut-off value of 2, set for the upper wall, which have a non-zero value for the potential. Switching off colvars has the MISC term go to zero, which confirms the MISC term is entirely from the harmonic walls restraint potential. Please help me understand why some values, within the harmonic walls have a non-zero value. 
> 
> Thanks for your help.
> 
> Following, is my colvars script: 
> 
> colvars on
> cv config "
> colvarsTrajFrequency 1
> colvarsRestartFrequency 1000
> colvar {
>     name d1
>     distance {
> group1 { 
>    dummyAtom (-4.317, -0.383, 0.141)
> }
> group2 { 
>    atomNumbersRange {1-17} 
> }
>     }
> }
> "
> 
> cv config "
> colvarsTrajFrequency 1
> colvarsRestartFrequency 1000
> colvar {
>     name d2
>     distance {
> group1 { 
>    dummyAtom (0.360, 0.03, 0.405)
> }
> group2 { 
>    atomNumbersRange {18-27} 
> }
>     }
> }
> "
> 
> cv config "
> harmonicWalls {
>     name hw1
>     colvars d1
>     lowerWalls -2.0
>     upperWalls 2.0
>     lowerWallConstant 2.0
>     upperWallConstant 2.0
> }
> "
> 
> cv config "
> harmonicWalls {
>     name hw2
>     colvars d2
>     lowerWalls -2.0
>     upperWalls 2.0
>     lowerWallConstant 2.0
>     upperWallConstant 2.0
> }
> "
> 
> Best Regards,
> Daipayan
> 
> ---------- Forwarded message ---------
> From: Daipayan Sarkar <sdaipayan_at_gmail.com <mailto:sdaipayan_at_gmail.com>>
> Date: Wed, Feb 5, 2020 at 2:34 PM
> Subject: Re: namd-l: Flat-bottom harmonic position restraints using colvars
> To: Giacomo Fiorin <giacomo.fiorin_at_gmail.com <mailto:giacomo.fiorin_at_gmail.com>>
> Cc: NAMD list <namd-l_at_ks.uiuc.edu <mailto:namd-l_at_ks.uiuc.edu>>, <dhardy_at_ks.uiuc.edu <mailto:dhardy_at_ks.uiuc.edu>>
> 
> 
> Hello Giacomo, 
> Many thanks for your quick response and sharing the link. After checking, harmonic wall potential is available through extrabonds. The NAMD documentation takes me to v2.13, where the usage of this new feature is not listed (standard harmonic potential for extrabonds). 
> 
> Can you or David (copying him here) help me with the usage of this feature, especially for the part "BondValue now has an additional parameter representing the upper wall."
> 
> Thank you again, 
> Daipayan 
> 
> On Wed, Feb 5, 2020 at 1:51 PM Giacomo Fiorin <giacomo.fiorin_at_gmail.com <mailto:giacomo.fiorin_at_gmail.com>> wrote:
> Hello Daipayan, those would work on the collective variables (i.e. one or a few numbers for the entire system) as opposed to an entire set of Cartesian coordinates.  And in any case, it would be quite inefficient to run a multitude of restraints via Colvars if they d don't need a collective operation.
> 
> Check out the nightly build, because a similar feature for Cartesian coordinates has just been added:
> https://charm.cs.illinois.edu/gerrit/c/namd/+/5316 <https://charm.cs.illinois.edu/gerrit/c/namd/+/5316>
> On Wed, Feb 5, 2020 at 1:29 PM Daipayan Sarkar <sdaipayan_at_gmail.com <mailto:sdaipayan_at_gmail.com>> wrote:
> Hello, 
> 
> My objective is to apply flat-bottom harmonic position restraints to the C-alpha (CA) atoms. To achieve this, I am using the colvars harmonicWalls. I have a question regarding setting the values of lower and upperWalls, with respect to the reference CA positions. 
> 
> Typically, a regular harmonic position restraint is applied, where a reference PDB is provided as an input to the parameter consref. My question, is can we pass a reference pdb to colvars and set the upper and lower bounds with respect to the selection? 
> 
> Following is my colvar definition. Any help is greatly appreciated.
> 
> #########################
> colvars on
> colvarsTrajFrequency 100
> 
> cv config "
> colvar {
>     name calpha
>     group1 {
>     psfSegID AP1
>     atomNameResidueRange CA 1-76
>     }
> }
> "
> 
> cv config "
> harmonicWalls {
>     name fb_restraint
>     colvars calpha
>     lowerWalls         #add_lower_bound_from_ref_CA_position
>     upperWalls        #add_upper_bound_from_ref_CA_position
>     lowerWallConstant 10.0
>     upperWallConstant 10.0
> }
> "
> #########################
> 
> 
> Thanks,
> Daipayan
> 
> 
> -- 
> Giacomo Fiorin
> Associate Professor of Research, Temple University, Philadelphia, PA
> Research collaborator, National Institutes of Health, Bethesda, MD
> http://goo.gl/Q3TBQU <http://goo.gl/Q3TBQU>
> https://github.com/giacomofiorin <https://github.com/giacomofiorin>
> 
> 
> -- 
> Giacomo Fiorin
> Associate Professor of Research, Temple University, Philadelphia, PA
> Research collaborator, National Institutes of Health, Bethesda, MD
> http://goo.gl/Q3TBQU <http://goo.gl/Q3TBQU>
> https://github.com/giacomofiorin <https://github.com/giacomofiorin>

This archive was generated by hypermail 2.1.6 : Thu Dec 31 2020 - 23:17:12 CST